Say No to Email Taxes

Today's new tax proposal comes from the British Prospect magazine:

Internet service providers (ISPs) have proposed price mechanisms to control it, but users objected. The time has come for a public sector remedy: a tax, perhaps no more than 2p, or 3c, on every email sent.

There's something strange about this argument. Here we have someone admitting that public opposition killed the private proposal to charge for email. Presumably this is because they thought the expense of paying an email charge was greater than the cost of dealing with spam. What does government involvement change about that calculation? Nothing. So why should the result be any different?

I also get incredibly annoyed at this argument:

How much would it cost? An average employee might send 100 emails a day. At 2p or 3c, the tax would be £2 or $3, less than a large caramel macchiato.

It's very common to compare some tax or cost to a cup of coffee a day, as if to make the point that it's completely trivial. But it isn't at all. For someone who works a full year, that's $750. Sure it's a 'cup of coffee', but guess what? I love vanilla lattes and almost never drink them because I can't afford them. Now you're going to tell me that emails will become as unaffordable to me as coffee. Great.

Then there's what the author wants to do with the revenue:

Above all, an email tax could safeguard the future of the internet itself. Peer-to-peer data transfers, video streaming and voice services like Skype demand ever greater bandwidth. When new capacity is needed, part of the tax proceeds could be used for investment.

Now this makes no sense, right? Suppose, as I have no reason to doubt, Skype and video streaming are what are causing internet bandwidth problems. Even if you accept this is a reason for government investment (rather than, say, abolishing net neutrality regulation), wouldn't you want to fund this investment from a tax on Skype, which is causing the problem in the first place?

As a libertarian-leaning fellow, I'm temperamentally opposed to new taxes in general, but this seems like an uncommonly bad proposal, even among tax hike initiatives.

Share this

Just say no

In the minds of some people, there is no problem which another tax won't fix.

I appreciate the way you took the "coffee comparison" and destroyed it. That one has bugged me for quite some time, as I too cannot afford that daily "designer" coffee. It bugs me to no end.

Thinking this through a little - $750 multiplied by numbers of users would buy some pretty awesome spam filtering, if that is what people wanted. A free market solution, when the free market really wants it, will kill this problem, and at less cost than the "public" solution.

Fully agreed: this is beyond

Fully agreed: this is beyond retarded.

I was just thinking about this thing yesterday: the obvious solution is to have an email account, where people need to send me a cent along with every email they send me. No third party runs off with the bounty.

The problem is implementational though: perhaps nobody could be bothered creating such a financially tied account as well. Itd have to be introduced by a big player, like google.

You could create an account with two folders: one inbox for all people whom had payed their dues, one inbox that would take anything else, with a mild spamfilter on top.

Itd completely eliminate spam; id be very interested. Most people seem to think it would be an outrage though. Needless to say, most people dont really think before they talk. If they do at all.

One time fee

What about a one-time fee of $1 to get a sender white-listed in your spam filter? Emails that don't make the cut are sent an auto-reply with payment instructions for your paypal account. Of course, you can white-list all your friends and business contacts and waive their fee when you set things up.

I think I read this idea originally on David Friedman's blog--maybe someone can find the exact post.

Yes, i forgot to mention:

Yes, i forgot to mention: whitelisting would make this all a lot easier.

But wouldnt it be vulnerable to adress faking?

Spoofed Address

It's trivial to spoof a "From" address, but the spammer would have to know an address in your whitelist, which isn't published. They could choose a popular one like "". For that matter, they could even pay you a buck to get in, then sell the paid "From" address along with your "To" address to millions of other spammers.

In both cases, you have reduced the number of available "From" addresses that get through from "infinite" (okay, limited by the number of characters and the character sets in an email address) to "finite, small, and known".

The first case still makes you vulnerable to phishing attacks, so you'd probably see the threat morph from wide advertising of anatomy-enhancing products to spear-phishing scams.

It's left as an exercise for the student to analyze the cost-benefit comparison of the two attacks...

legitimate mass email

Of course this completely fails to consider the costs to firms that legitimately send millions of emails. The company I work for sent approximately 135 million emails last year, which at 2 cents a pop would run up a tax bill of 2.7 million dollars.

Most of that email is absolutely essential to the way this company conducts business and is NOT spam. Obviously we would have to consider each email we send in terms of its revenue impact, and that would reduce our volume slightly, but the main effect would simply be a transfer of income from us to the state.

Not sure it's even possible

I'm not sure it's even possible to tax email in any effective way. First of all, what is email? Can it be defined and effectively taxed in a way that taxes just email without taxing, say, instant messaging, or file sharing, or web browsing? If there is a website where you and your friends have an account and you can leave messages for them at that website which they can then pick up, is that email? And can it be targeted for a tax without also taxing all web browsing? Can such things be targeted for tax without requiring that the legislature reconvene every two weeks to update the definition of email?

Email is a way of using digital transfers. Analogy: you can tax, say, pencils and paper, and thus, indirectly, you can tax all the things that can be done with pencil and paper. But you can't specifically tax doodling, or writing limericks, while leaving other uses of pencil and paper untaxed. There are limits on what the government can actually tax.

The government might be able to tax a given email system, and maybe network effects will keep people on the taxed system, but as a matter of fact I don't think people are all that wedded to any given system. Many people I know happily juggle multiple email accounts and happily simultaneously use email-like systems which are not email, alongside regular email. Because of this, I don't think the network lock-in is that strong.

One line summary: I highly doubt Edward Gottesman has the first clue what he's talking about.

Edit: From the article:

From a practical point of view, such a tax is feasible. Whether you’re using a browser or a client-based email system, every email sent must have both a sender address and a recipient address—each in the form “someone@somewhere.” This makes all emails easily identifiable by ISPs, through which most private internet traffic is routed.

This is false. I receive and send email at gmail. Gmail is not an ISP. When I access gmail, the connection is encrypted. Therefore it is false that "this makes all emails easily identifiable by ISPs". None of my email - none whatsoever - is identifiable as such by my ISP, and this has been the case for years.

So, yes indeed, Edward Gottesman does not know what he is talking about.

Gmail is not an ISP. When I

Gmail is not an ISP. When I access gmail, the connection is encrypted. Therefore it is false that "this makes all emails easily identifiable by ISPs". None of my email - none whatsoever - is identifiable as such by my ISP, and this has been the case for years.

Thanks for pointing that out; I had wondered about that myself.

There will never be an honest, secret ballot

in an election involving millions of votes. Never was.

I suggest that in a national election that the county voting lists and all ballots be published as down loadable data. Every person can then check for dead neighbors voting and make their own tally. A simple BASIC program or spread sheet should work.

Every candidate list should include "none of the above." IF NOTA wins then the election should start over from scratch.

Better, we adopt a parliamentary system.

Better, we keep the ballot

Better, we keep the ballot dirty shady and not trustable.

tax the largest users and leave everyone else alone

How about give everyone a "free" 100 emails a day. If you send more than 100 emails a day or 1500 a month, you have to apply for a license - describing your purpose of sending mass emails, and the legitimacy of it (that it is not spam) and for a one time fee of $50 a year you get to send unlimited emails. If later, these emails are found in fact to be spam, your license is revoked. Everyone else is blacklisted after sending more than 100 a day, automatically.

Instead of going by "from" address, you could use the origin IP in the headers of the email to identify senders. I'm not exactly sure how that works, but I know if I use an email application that uses my own PC as the origin for outgoing mail, its most likely going to end up in other people's trash because it ends up outside the expected channels of properly sent emails.

This gives 99% of email users a free ride without having to touch a thing. Large businesses sending tons of emails pay a one time small fee. The monies collected could go towards policing the email protocol and arbitrating instances where someone claims spam is going on (and deciding if it really is or not).

This wouldn't stop someone from using, say hotmail, for spam ... but the large web email providers already have functions in place to catch people using their services for spam.


All of these proposals are defeated by botnets. Stop trying social engineering, it *doesn't work*.