March 06, 2006
The economics of email
Last week I received the following email message. I bet some of you did, too.
How are you doing. Hope you have not forgotten me, I am Mr.MarlvinSomtochukwu. the man from Nigeria who contacted you some time ago to assist me secure the release of some money accrued from the over invoicing of a contract that was awarded by my government through my ministry during the militry regime some years ago. Though you were not able to asist me conclude the transaction, I'm happy to inform you about my success in getting those funds transffered under the assistance and cooperation of a new partner from Paraguay.
Presently i'm in South Korea for investment projects with my own share of the total sum. meanwhile, ididn't forget your past efforts and attempts to assit me in transffering those funds, I made sure you are not leftout in the benefit of the transaction hence I kept aside for you a certified cashier cheque covering the sum of $250,000.00. (Two Hundred and Fifty Thousand United state Dollars).
Why, Mr. MarlvinSomtochukwu, how could I forget you? How often do you think it is that somebody offers me a sum of $250,000.00 (Two Hundred and Fifty Thousand United state Dollars)? Not more than once every 45 minutes, I'm guessing.
I used to think that this limitless supply of Mr.MarlvinSomtochukwu's was the result of a strange folk superstition rampant in Nigeria that if you spun such a yarn to a stupid American, they would give you all their money. But apparently every now and then there is some dope who actually falls for it. The most recent would-be gazillionaire appears to be Dr. Louis A. Gottschalk, whom the LA Times describes as a "world-renowned psychiatrist" who "gave as much as $3 million over a 10-year period in response to an Internet plea that promised the doctor a generous cut of a huge sum of cash trapped in African bank accounts in exchange for money advances."
Hmmm, a psychiatrist, you say? Isn't that somebody who is supposed to be an expert about why people sometimes do crazy things?
Appeals on behalf of these millions of dollars trapped in Nigerian banks are just a small part of the several hundred completely useless email messages that I receive every day. Although every now and then they are a little amusing, for the most part I regard these as simply a curse, requiring me to at least glance at the titles of the messages trapped by my spam filter, and occasionally causing me to completely ignore legitimate correspondence.
From an economic point of view, the nature of the spam problem is that it costs far less for the producer to send one of these messages than it costs the recipient to process it. There is for this reason way too much of this "good" being produced, a classic case of a market externality.
The classic solution to this kind of externality is to impose a tax on the producer. If it cost Mr. MarlvinSomtochukwu even 1/10 of a cent to send this message, I'm sure he wouldn't think it worth his while to bother sending it.
When AOL proposed a fee that bulk emailers could pay to be exempt from company spam filters, there was a huge cry of protest from surprisingly diverse groups. And yet, a little reflection should convince you that being charged a fee in order to send an email message is exactly the correct solution. Ideally, each of us could set a price for which we would need to be compensated in order to receive an email message-- personally I'd charge maybe one cent. The email service provider could add their own fee-- say another half cent more-- and a sender would decide whether they wanted to fork out the 1.5 cents to get the message through. A given bulk emailer could choose to send its message to all the folks who charge less than 2 cents per pop, for example.
Even if I didn't get the penny myself, I'd still be better off under such a system. If it isn't worth 2 cents to you to want to send your message, it isn't worth my time to bother glancing at the message title.
Who's going to be deterred from sending an email message if they have to pay a penny to send it? Well, the people who currently are sending their messages out to 25 million people. By my reckoning, they'd think twice about paying $250,000.00 (Two Hundred and Fifty Thousand United state Dollars) just in order to clutter up everybody's email inbox.
Posted by James Hamilton at March 6, 2006 11:55 AMdigg this | reddit
I like this idea, but I would focus on the concept that the recipient of the email is the one who gets the cash. He is the one who pays most of the cost of manually scanning his spam folder for misfiled email or, worse, manually scanning his email folder and deleting tons of spam. The ISP merely has a larger volume of email traffic to send, but compared to other Internet protocols like video and file downloads, I doubt that Internet spam is a major cost for them. Besides, the ISP is paid already, by the customer. It is the customer who is being uncompensated for dealing with spam.
Solutions like AOL's where the company gets the profits and end users get nothing are the wrong direction to go. The compensation is not going to the party who is most directly impacted by the spam, so the economic effects will not be as beneficial as they might be.
Posted by: Hal at March 6, 2006 01:22 PM
The end users get nothing?
He just explained that by making it cost something to bulk mail, less bulk mail gets sent, meaning the end users get a more pleasant experience.
It's rather like a cover charge to keep the riff-raff out.
Posted by: Ben Cremeens at March 6, 2006 03:48 PM
I'll divide this comment into three parts
- a comment on the practicality of the above proposition.
- my thoughts about the concept of bulk mailing
- a general note on the above 419 scam (thats what it is called)
I think there is an assumption here that monitoring and controlling bulk mailing is something easy to do. Perhaps this is possible just within AOL microcosm but in general, this isn't an easy problem to solve. One can fake addresses, or even information about the path taken by emails, the source it came from. It is not hard to automate ways of generating such spam emails with relatively identical content but which can slip through a monitoring system designed to catch bulk mails. But arguably the ones who land up sending the bulk emails are not necessarily the smart folks.
That said, I think your point touches upon a more general problem. I am all for the idea of adding a barrier to sending bulk emails or in general increasing the cost to the producer for anything where the cost of sending unsolicited "stuff" to a consumer is negligible compared to the cost to the consumer for having to process the unwanted "stuff' (and filter it out from the relevant stuff). This includes a lot of the junk (snail) mail I get in my mail box as well as marketing calls which till some time back had reached really painful limits.
Lastly, just a note on the Nigeria 419 scam. As pointed out in the post, there is always someone or the other who falls for it. There is a song made based on this whole concept. The world tech podcast has done a small segment on it. It includes the song "I go chop your dollar" which became a hit on the Nigerian radio stations. Its an interesting piece.
Posted by: Vikram Asrani at March 6, 2006 04:07 PM
Perhaps they could tag the message with the charge and users could collect for reading it. If they really want people to read it they can make it 10, 20, 30 cents or more. A budget advertising model.
Once you find a sucker, there is nothing they won't fall for, so now we can all beggar Dr. Gottschalk to fork it over. (Related to the department stores of the same name?)
Posted by: Lord at March 6, 2006 05:03 PM
James, a propos of the original email that inspired your post have you seen this web page?
Posted by: Dave Schuler at March 6, 2006 06:23 PM
The problem with AOL's proposal isn't that they're charging for email--in fact they aren't. They're charging for the guarantee that the email will end up in the user's inbox, and not a junk mail folder. Every AOL user will still get *all* the junk mail they currently get, plus an additional amount that gets directed straight to their inbox because some company paid for it. This certainly is not a benefit to AOL's users.
(The only way this could benefit AOL users is if they've been having a problem with false positives going to the junk mail. But I almost never hear people complain about that. That may just be because no one ever knows what they're missing. ;) )
But JH isn't exactly talking about AOL's proposal. I don't like these kinds of proposals, but for those promoting them, I would suggest that one could accomplish the same purpose in a more egalitarian manner: Charge X to people to send an email, and credit them X when they receive an email. If you send as much as you receive, you won't pay anything. This still keeps the spammer at bay, but diminshes the distortion between two friends emailing each other.
On a technical standpoint, though, charging for email is only possible if you fragment the Internet. I run Linux at home and every once in a while, just because I can, I turn on my email server and send out email directly from my home computer. I can't imagine having a system of needing to pay to send an email to AOL.com that would still allow me to send email from my home; AOL, Google, MS simply wouldn't find it worth their time to set up an account for me. Ultimately, the only people who could send and receive email are the ones large enough that it's worth including them in a payment network. You can see things unravel from there--China refuses to play along, so no more email between the U.S. and China. Microsoft charges different amounts to different senders.....
(Oh, and I find it ironic for this post that the blog software requires a valid email to post, when publicly posted email addresses is the best way for spammers to get your address. Forgive me if I jiggered my email a bit...)
Posted by: mlmitton at March 6, 2006 06:26 PM
Dave, that's too funny!
mlmitton, right on about the irony. That's another one of those details about MovableType that I should try to figure out how to master in order to avoid exactly the problem you mention.
Posted by: JDH at March 6, 2006 06:55 PM
And people think car salesman are sleazy!
Posted by: Rick at March 7, 2006 05:52 AM
Well, it costs money now to send junk (snail)mail, yet I still get junk mail. Why? Because the junk mailers pay the post office real money to advertise. I understand that the postal service even gives discounts to bulk mailers.
Since IMO there's no way that you and I are going to be paid for getting email, our ISPs probably stand to collect whatever money is involved. So while charging for email might stop Nigerian or Ukrainian spammers, it might also encourage ISPs to see junk email as a profit opportunity. I'd expect to see fewer shot-in-the-dark spams, but more of the targeted advertising spams that are so prevalent in our physical mailboxes.
Posted by: Donal at March 7, 2006 06:42 AM
Two quick points - one, some bulk email is actually welcome (many hobby groups communicate with email listservers) and two, the enforcement mechanisisms carry some very serious external costs.
Posted by: Mike at March 7, 2006 07:08 AM
The deeper question is, why do seemingly inefficient outcomes (in the form of the Nigerian e-mail) persist when the Coase theorem says they shouldn't? Five years ago, Charlie Kahn, Jamie McAndrews & I wrote a paper (available at frbatlanta.org) where we discuss why not. The answer is that Coasian bargaining isn't always effective when certain parties to a transaction(e.g., e-mail scam artists) can undertake hidden actions unobservable by other parties (e.g., spam). With this barrier to Coasian bargaining, we would expect reassignment of property rights to matter.
Posted by: Will at March 7, 2006 11:05 AM
Micropayments seem a good way to combat email spam, much more effective than a filter. People have become accustomed to free email, so it probably makes most sense to make it revenue-neutral. Users could be allowed to authorize certain email addresses, so as to receive newsletters. And for users who email back and forth, if I send you a penny and you send it right back, we're all good.
Posted by: Ben at March 7, 2006 11:33 AM
You are typically a desk economic prof. Reality is diffrent.
It cost tons of money to send snail mail to my home. Yet every day my mailbox is filled with ads of all kinds. Even more then I get in my email box.
Spam is an easy problem to solve. We only need to change the email protocol a little bid. The open source groups propose this.
Further more you take away the free character of email en the internet.
Posted by: Freedom at March 8, 2006 01:12 AM
Techno-solutions don't always work as intended over "public access" networks, however. E.g., even a desk econ prof. might realize that the introduction of credit cards, a sort of technologically superior check, helped retailers control the problem of bad checks written at the point of sale. Yet by virtue of their success, credit cards have also given us the banking equivalent of spam, aka identity theft. The desk economist might further point out that spam prevention is a nonrival good, and the provision of such goods is always a tricky business, both online & off.
Posted by: Will at March 8, 2006 04:44 AM
While the economics of this idea is sound, it is technically not an easy thing to solve.
The problem today is that a lot of spam is NOT sent by one bulk email server, but rather surreptitiously by bot-nets. (more later).
I receive every day in my email box the same exact email, advertizing pharmaceutical products, but the sender is always new, but real.
And now to how this is done: the sender has access to a large number of hacked PCs which he can control, and cause to be the sender of email.
And this points us to the weakness of all email cost schemes: It's not the real sender that gets to foot the bill but some other poor soul.
Charging for email will only ripp of two persons: the involontary sender and the involontary receiver.
You could say that this will teach the unprotected to protect themselves, but there will always be a majority of users who don't have the wherewithall to patch their PC, and to mount a firewall and to continously keep this up - being only 1 hour late with this, opens you up to the possibility of being hacked - Hacking is now automated.
When I'm looking at my firewall logs I can see that an attempt to hack me is done about once every hour.
There are statistics showing that an unpatched system connected to the Internet lasts an average of 26 minutes before being compromized.
Posted by: pfalk at March 8, 2006 09:56 PM
This seems related to to the practice of TV screens everywhere in public, particularly airports. (I just flew cross-country on business.)
I think that some content providers PAY the terminal operators to use their feeds so that they can boost of higher audience and so collect higher ad rates.
I know that in my office, the company installed TV screens fixed to the company-owned network. Most employees find the TVs annoying and a disruption but who asked for TVs in the breakroom?
Another case is the airlines intruding on the passengers with TV screens from which there is no escape.
The similarity is that they should PAY me, as a captive audience, to be exposed to their advertisements.
Posted by: Joseph Somsel at March 9, 2006 03:08 PM