Slashdot.com, April 29, 2009
Frequent Slashdot contributor Bennett Haselton writes with his take on the recent Time Warner Cable fiasco: “Net Neutrality crusaders at FreePress.net recently called attention to Time Warner’s plan (later rescinded) to impose fines on users for going over bandwidth limits. I agree generally, but I think this is easily confused with the reasoning in favor of Net Neutrality, and it’s important to keep the arguments separate.” Read on for the rest of Bennett’s thoughts.
On April 13th I received an e-mail from FreePress.net, one of the organizations that led the fight in favor of Net Neutrality:
Just as we’re suffering economically, Time Warner Cable is trying to squeeze us even further, forcing millions of customers to pay steep fees for exceeding an absurdly low monthly limit on Internet use. […] The company’s scheme would cost customers $15 per month for one gigabyte — the equivalent of one 30-minute HD television show — with a penalty fee of $2 for every additional gigabyte over the limit.
Later, FreePress.net triumphantly announced that Time Warner had reversed their position. Now, I would appear to have painted myself into a corner on this issue, because I wrote in an editorial two years ago arguing in favor of Net Neutrality:
[Net Neutrality is] not about how much a service costs, but about the ethics of double-billing for it. […] If vastly more people start trying to stream CNN over the Internet 24/7, and fully using the services that ISPs have “only been pretending to sell,” as Brad Templeton put it, then ISPs may have to charge more for users who consume too much bandwidth, encouraging people to stay at today’s average levels by rationing themselves and perhaps watching 24 on their $5,000 TV sets sometimes instead of downloading it off of BitTorrent to their laptop every week because it makes them feel like a haX0r. Much as we all love our unmetered connections, it wouldn’t be a violation of Net Neutrality for ISPs to charge users for bandwidth hogging, to keep everyone from going too far above today’s levels.
And yet, even after writing those words, I still think there is an argument against letting ISPs impose bandwidth fines, at least under some conditions. However, I think the argument is completely separate from the argument in favor of Net Neutrality, so it’s important to derive both of them independently of each other.
I would try to make both arguments by deriving the conclusions from first principles. This might seem pedantic at times, but I think it’s helpful to have a precise mathematical-style “proof” of why a conclusion follows from its premises, because then you can see how changing one premise would change the conclusion.
To me the simplest argument in favor of Net Neutrality follows from three assumptions. You don’t have to agree with the assumptions, but I think that all three of them are obvious because the opposite would be untenable.
1. An ISP that blocks (or slows access to) certain websites is defrauding its users UNLESS either (a) the ISP has made its users aware of the filtering, or (b) it’s overwhelmingly clear that the filtering protects the users or improves their experience (so more experienced users would assume it is taking place anyway). If your ISP has told you that they’re selling “Internet access” but they’re silently blocking some Web sites, then this is straightforward. You’re paying for one thing, and the ISP is selling you something else that is inferior. In the incident that I wrote about, ISPs like Rogers.com that used AboveNet as their upstream provider, were actually blocking their subscribers from reaching certain websites, even though their customers thought they were getting unfiltered Internet access. Now if the ISP advertises that its Internet connections are filtered, as some “family friendly” providers do — so that virtually all users knew about the filtering — then this would not be a violation of Net Neutrality. And if the ISP is blocking mail from actual spam sources, then this is something that protects users and improves their experience, and so is usually not considered a violation of Net Neutrality either. But if the ISP is silently blocking access to Web sites, or blocking mail from servers that are not sending spam but simply because the ISP owner has a political disagreement with those server owners, then that would violate this principle.
2. “Make its customers aware” means just that — make its customers aware — and not bury something in the Terms of Service. Imagine if the opposite principle were accepted — that websites and software vendors could do anything they wanted as long as they put the right disclaimer in the 23rd paragraph of their site’s or program’s “Terms of Service” that nobody reads. Scam artists’ eyes everywhere would light up with dollar signs thinking of the possibilities: Create a popular program and get people to install it, while putting a clause deep in the TOS that permits them to remotely take over your computer after you’ve installed their software! Or for a real-world example, Yahoo! once tried to amend the GeoCities Terms of Service to give Yahoo! the copyright on any content uploaded by their users. Yahoo! reversed itself after a public backlash, but even if they hadn’t, it would have been good public policy for a court to say that Yahoo!’s copyright claim on their users’ content was invalid. You can, of course, strengthen your legal rights by putting the right language in your Terms of Service, but it would mean total chaos if companies could bury “gotchas” in your TOS that are wildly contrary to what users are reasonably likely to assume.
3. If company A sells something to company B which company B then re-sells to the public, but company B almost certainly cannot resell the good without committing fraud as outlined above, then company A is complicit in the fraud as well. Some of AboveNet’s defenders argued that they mostly sold Internet connectivity to ISPs, not to the public, and the ISPs knew that the connections were filtered. Even assuming this were true, the ISPs still would not be able to re-sell the service to the public without representing it as “regular Internet access” — nobody would pay full price for a broken or degraded connection when a competitor could offer a regular connection for the same price.
So, an ISP that blocks or degrades access to certain Web sites, when users think they are getting full unfettered Internet access, is cheating customers (or, in the case of a backbone provider, complicit in the downstream ISPs cheating their customers) in violation of the principles of Net Neutrality. QED. I would tentatively call these assumptions airtight; at least, I cannot think of any corporate behavior that violates one or more of these principles and should be allowed under good public policy.
By contrast, the argument against Net Neutrality — that the free market will ensure that ISPs provide effective service without the need for government regulation — relies on assumptions that might sound reasonable, but have loopholes, and the loopholes are precisely where Net Neutrality violations can slip through. An anti-Net-Neutrality editorial by Sonia Arrison, for example, argued that “consumers would never stand for blocked Web sites.” However, in the case of AboveNet’s filtering, downstream users did of course “stand for it,” because they didn’t know about it, and the natural assumption, when the user sees a website not responding, is to think that the site is down, not that their provider blocked it.
But the argument against bandwidth fines is different. While “broken” Internet access could never be sold to the public without some sort of misrepresentation, it is conceivable that people would still pay for Internet access even if the price were $15 for the first 1 GB and $2 per GB after that. However, it would still be good public policy to prohibit two variants of this scheme: (a) ISPs silently racking up charges, scummy-cell-phone-company style, against users who may not realize what charges they’re incurring, and then shocking them with overage bills at the end of the month; and (b) ISPs charging draconian bandwidth fines in cases where they have a monopoly, or near-monopoly, on users’ Internet access options.
Prohibiting “shock” overage bills essentially follows from principles #1 and #2 above — users should know what they’re getting, and sneaking something into the fine print doesn’t count. If someone is approaching their bandwidth limit, and is on track to run over (and incur a lot of charges) before the end of the month, it wouldn’t be too much trouble to send them an e-mail or an automated (or live) phone call to warn the user what’s going on. If the ISP objects that this would cost them too much, I’d say I’ll happily pay $1 for the trouble of them placing a call to my house if it saved me $20 in surprise overages.
Prohibiting bandwidth fines in the case of monopoly situations simply follows from the principle that without competition, the bandwidth overage fees are likely to be much higher than they would be in a competitive market. It may not be the motivation of the ISP simply to make as much money as possible; perhaps they want to discourage high-bandwidth usage for other reasons. As FreePress.net theorized about the proposed Time Warner bandwidth surcharges: “This trick is designed to make customers think twice before switching off their cable TV and finding the shows they want online.” But whether it’s to squeeze subscribers for extra money or to stop them from streaming content from the Internet, either way, the plan could not be sustainable if users can find higher bandwidth at a lower cost from other providers. For most of its subscribers, Time Warner doesn’t have a pure monopoly — in some areas, you can get only one cable Internet provider and only one DSL provider, but the two still compete with each other to provide “Internet access,” and other areas have a choice between cable providers. However, in a situation with only a small number of competitors, companies can still keep prices higher than they would in a purely competitive market, because there are fewer chances for an upstart competitor to find ways to provide a more efficient service at a lower cost.
What if neither of these conditions were true? If an ISP actually did make sure that its subscribers knew about the bandwidth limits, and users got warnings if they were approaching those limits, and there were enough competing providers to ensure real competition, then in that situation it would be harder to make an argument against the bandwidth surcharges. (Admittedly, it may be something of an academic point, because there are so few situations where there are “enough competing providers” to guarantee healthy competition.)
But it’s important to keep the arguments for Net Neutrality separate from the arguments against bandwidth surcharges. Bandwidth fines are bad mainly when there are few competing providers, because it will be hard for users to get a better deal somewhere else, and providers like Time Warner may have a vested interest in keeping users’ bandwidth limits low to keep them glued to their TV. Violations of Net Neutrality are bad regardless of whether there are few or many competing providers, because users cannot avoid the harm if they’re unlikely to discover what’s happening in the first place.
Bandwidth Fines Bad, But Not Net Neutrality Issue
Pacific Research Institute
Slashdot.com, April 29, 2009
Frequent Slashdot contributor Bennett Haselton writes with his take on the recent Time Warner Cable fiasco: “Net Neutrality crusaders at FreePress.net recently called attention to Time Warner’s plan (later rescinded) to impose fines on users for going over bandwidth limits. I agree generally, but I think this is easily confused with the reasoning in favor of Net Neutrality, and it’s important to keep the arguments separate.” Read on for the rest of Bennett’s thoughts.
On April 13th I received an e-mail from FreePress.net, one of the organizations that led the fight in favor of Net Neutrality:
Just as we’re suffering economically, Time Warner Cable is trying to squeeze us even further, forcing millions of customers to pay steep fees for exceeding an absurdly low monthly limit on Internet use. […] The company’s scheme would cost customers $15 per month for one gigabyte — the equivalent of one 30-minute HD television show — with a penalty fee of $2 for every additional gigabyte over the limit.
Later, FreePress.net triumphantly announced that Time Warner had reversed their position. Now, I would appear to have painted myself into a corner on this issue, because I wrote in an editorial two years ago arguing in favor of Net Neutrality:
[Net Neutrality is] not about how much a service costs, but about the ethics of double-billing for it. […] If vastly more people start trying to stream CNN over the Internet 24/7, and fully using the services that ISPs have “only been pretending to sell,” as Brad Templeton put it, then ISPs may have to charge more for users who consume too much bandwidth, encouraging people to stay at today’s average levels by rationing themselves and perhaps watching 24 on their $5,000 TV sets sometimes instead of downloading it off of BitTorrent to their laptop every week because it makes them feel like a haX0r. Much as we all love our unmetered connections, it wouldn’t be a violation of Net Neutrality for ISPs to charge users for bandwidth hogging, to keep everyone from going too far above today’s levels.
And yet, even after writing those words, I still think there is an argument against letting ISPs impose bandwidth fines, at least under some conditions. However, I think the argument is completely separate from the argument in favor of Net Neutrality, so it’s important to derive both of them independently of each other.
I would try to make both arguments by deriving the conclusions from first principles. This might seem pedantic at times, but I think it’s helpful to have a precise mathematical-style “proof” of why a conclusion follows from its premises, because then you can see how changing one premise would change the conclusion.
To me the simplest argument in favor of Net Neutrality follows from three assumptions. You don’t have to agree with the assumptions, but I think that all three of them are obvious because the opposite would be untenable.
1. An ISP that blocks (or slows access to) certain websites is defrauding its users UNLESS either (a) the ISP has made its users aware of the filtering, or (b) it’s overwhelmingly clear that the filtering protects the users or improves their experience (so more experienced users would assume it is taking place anyway). If your ISP has told you that they’re selling “Internet access” but they’re silently blocking some Web sites, then this is straightforward. You’re paying for one thing, and the ISP is selling you something else that is inferior. In the incident that I wrote about, ISPs like Rogers.com that used AboveNet as their upstream provider, were actually blocking their subscribers from reaching certain websites, even though their customers thought they were getting unfiltered Internet access. Now if the ISP advertises that its Internet connections are filtered, as some “family friendly” providers do — so that virtually all users knew about the filtering — then this would not be a violation of Net Neutrality. And if the ISP is blocking mail from actual spam sources, then this is something that protects users and improves their experience, and so is usually not considered a violation of Net Neutrality either. But if the ISP is silently blocking access to Web sites, or blocking mail from servers that are not sending spam but simply because the ISP owner has a political disagreement with those server owners, then that would violate this principle.
2. “Make its customers aware” means just that — make its customers aware — and not bury something in the Terms of Service. Imagine if the opposite principle were accepted — that websites and software vendors could do anything they wanted as long as they put the right disclaimer in the 23rd paragraph of their site’s or program’s “Terms of Service” that nobody reads. Scam artists’ eyes everywhere would light up with dollar signs thinking of the possibilities: Create a popular program and get people to install it, while putting a clause deep in the TOS that permits them to remotely take over your computer after you’ve installed their software! Or for a real-world example, Yahoo! once tried to amend the GeoCities Terms of Service to give Yahoo! the copyright on any content uploaded by their users. Yahoo! reversed itself after a public backlash, but even if they hadn’t, it would have been good public policy for a court to say that Yahoo!’s copyright claim on their users’ content was invalid. You can, of course, strengthen your legal rights by putting the right language in your Terms of Service, but it would mean total chaos if companies could bury “gotchas” in your TOS that are wildly contrary to what users are reasonably likely to assume.
3. If company A sells something to company B which company B then re-sells to the public, but company B almost certainly cannot resell the good without committing fraud as outlined above, then company A is complicit in the fraud as well. Some of AboveNet’s defenders argued that they mostly sold Internet connectivity to ISPs, not to the public, and the ISPs knew that the connections were filtered. Even assuming this were true, the ISPs still would not be able to re-sell the service to the public without representing it as “regular Internet access” — nobody would pay full price for a broken or degraded connection when a competitor could offer a regular connection for the same price.
So, an ISP that blocks or degrades access to certain Web sites, when users think they are getting full unfettered Internet access, is cheating customers (or, in the case of a backbone provider, complicit in the downstream ISPs cheating their customers) in violation of the principles of Net Neutrality. QED. I would tentatively call these assumptions airtight; at least, I cannot think of any corporate behavior that violates one or more of these principles and should be allowed under good public policy.
By contrast, the argument against Net Neutrality — that the free market will ensure that ISPs provide effective service without the need for government regulation — relies on assumptions that might sound reasonable, but have loopholes, and the loopholes are precisely where Net Neutrality violations can slip through. An anti-Net-Neutrality editorial by Sonia Arrison, for example, argued that “consumers would never stand for blocked Web sites.” However, in the case of AboveNet’s filtering, downstream users did of course “stand for it,” because they didn’t know about it, and the natural assumption, when the user sees a website not responding, is to think that the site is down, not that their provider blocked it.
But the argument against bandwidth fines is different. While “broken” Internet access could never be sold to the public without some sort of misrepresentation, it is conceivable that people would still pay for Internet access even if the price were $15 for the first 1 GB and $2 per GB after that. However, it would still be good public policy to prohibit two variants of this scheme: (a) ISPs silently racking up charges, scummy-cell-phone-company style, against users who may not realize what charges they’re incurring, and then shocking them with overage bills at the end of the month; and (b) ISPs charging draconian bandwidth fines in cases where they have a monopoly, or near-monopoly, on users’ Internet access options.
Prohibiting “shock” overage bills essentially follows from principles #1 and #2 above — users should know what they’re getting, and sneaking something into the fine print doesn’t count. If someone is approaching their bandwidth limit, and is on track to run over (and incur a lot of charges) before the end of the month, it wouldn’t be too much trouble to send them an e-mail or an automated (or live) phone call to warn the user what’s going on. If the ISP objects that this would cost them too much, I’d say I’ll happily pay $1 for the trouble of them placing a call to my house if it saved me $20 in surprise overages.
Prohibiting bandwidth fines in the case of monopoly situations simply follows from the principle that without competition, the bandwidth overage fees are likely to be much higher than they would be in a competitive market. It may not be the motivation of the ISP simply to make as much money as possible; perhaps they want to discourage high-bandwidth usage for other reasons. As FreePress.net theorized about the proposed Time Warner bandwidth surcharges: “This trick is designed to make customers think twice before switching off their cable TV and finding the shows they want online.” But whether it’s to squeeze subscribers for extra money or to stop them from streaming content from the Internet, either way, the plan could not be sustainable if users can find higher bandwidth at a lower cost from other providers. For most of its subscribers, Time Warner doesn’t have a pure monopoly — in some areas, you can get only one cable Internet provider and only one DSL provider, but the two still compete with each other to provide “Internet access,” and other areas have a choice between cable providers. However, in a situation with only a small number of competitors, companies can still keep prices higher than they would in a purely competitive market, because there are fewer chances for an upstart competitor to find ways to provide a more efficient service at a lower cost.
What if neither of these conditions were true? If an ISP actually did make sure that its subscribers knew about the bandwidth limits, and users got warnings if they were approaching those limits, and there were enough competing providers to ensure real competition, then in that situation it would be harder to make an argument against the bandwidth surcharges. (Admittedly, it may be something of an academic point, because there are so few situations where there are “enough competing providers” to guarantee healthy competition.)
But it’s important to keep the arguments for Net Neutrality separate from the arguments against bandwidth surcharges. Bandwidth fines are bad mainly when there are few competing providers, because it will be hard for users to get a better deal somewhere else, and providers like Time Warner may have a vested interest in keeping users’ bandwidth limits low to keep them glued to their TV. Violations of Net Neutrality are bad regardless of whether there are few or many competing providers, because users cannot avoid the harm if they’re unlikely to discover what’s happening in the first place.
Nothing contained in this blog is to be construed as necessarily reflecting the views of the Pacific Research Institute or as an attempt to thwart or aid the passage of any legislation.