The article, when read carefully, seems quite disingenuous, because given the way the points are carefully phrased to sound sensible while actually being wrong, it is hard to believe that the author actually believes the argument themselves.
Going through it point by point:
* Their first argument against is that it is hard to define network discrimination. But it really isn't, and I can define it in one sentence: routing network packets differently depending on anything other than header information specifically designed for routers to process (IP headers in the case of Internet Protocol), or routing outgoing traffic based on destination addresses with the intention of deliberately slowing traffic to some destinations but not others.
* 2 - to paraphrase "Internet access is not a right, so we shouldn't protect it". The reality is that so much human communication, including participation in democratic processes and a huge part of the economy happens online, and so it is reasonable to consider it a human right. The fact that it is relatively new doesn't change that.
* 3 - To paraphrase "Network discrimination can be good, because it lets operators:
Stop immoral practices like spamming or 'hacking'"
(presumably they mean cracking). That some people might do bad things not a good reason to censor the Internet. Favouring downloads over uploads for technical reasons - that is not really within the scope of network neutrality.
Conserve bandwidth - they give examples about usage caps, which are content neutral so outside of the scope of network neutrality. Specifically trying to slow certain applications, relative to another, is not necessary to limit bandwidth usage in a neutral way.
Promote brands - this is exactly why network neutrality is needed - network operators shouldn't be able to exploit their market position as gatekeeper to the Internet to favour one company over another.
Preventing 'theft' of service - except that if I pay for bandwidth, it is not theft to use it however I like, including allowing others to use it. Another good argument for network neutrality, not against.
Implementing a variety of network access charges to help to recover costs - ISPs can make money charging users for bandwidth in a non-discriminatory fashion, so there is no need for them to exploit their market position to charge people who aren't even connected to the network as well - and stopping them from doing that is a legitimate aim of network neutrality.
* 4 - to paraphrase "consumers might want more than a dumb pipe from their ISP, and if they go too far, the market will take care of it"; ISPs can offer services which are optional to consumers - that isn't a network neutrality issue, it only is when they become compulsory (e.g. you must use our nameservers) that it is. No consumer wants to be forced - at best they will tolerate it or it will make no difference.
* 5 - to paraphrase, "the market will sort it out". The market can't sort it out, because there are high barriers to entry, it is a finite market, and so the market is not infinitely efficient (and in some areas approaches a monopoly), and in practice there is little room for consumer choice influencing Internet markets except with regards to price, advertising, and to some extent reliability. While consumers benefit from network neutrality, very few are in a position to take this into account when choosing services.
* 6 - To paraphrase "ISPs should be free to put whatever they want in their contracts". The reality is that 'freedom of contract' is not fair when there is a huge difference in bargaining power, the markets can't sort it out for the reasons discussed above, and government intervention is appropriate.
* 7 - To paraphrase, "Network operators own their equipment and should be able to do what they like with it". That is like saying gun owners should be able to do what they like with their gun, even shoot people. It is reasonable to put in place laws against certain types of unreasonable exploitation of a market position, and property rights don't extend to using the property to break those reasonable laws. The argument here seems particularly disingenuous.
* 8 - To paraphrase, "network neutrality will discourage innovation and investment". The reality is that network operators can still profit from charging users for access - and that is enough to encourage investment. The main limit on ISPs profits is likely to be competition, and all ISPs are subject to the same laws - so no service that was profitable without network neutrality is likely to become unprofitable because of network neutrality.
* 9 - to paraphrase "introducing laws around the Internet will encourage players with a special interest to try to game the system". There is an obvious and simple solution to this one - transparency and openness in decision making. Gaming works both ways - gaming to not introduce network neutrality because it is in corporate interests anyway.
* 10 - to paraphrase - "putting in place Network Neutrality laws will start down a slippery slope towards government control of the network". Nothing specific to any network neutrality law makes it easier to change laws around the Internet later, so the argument is wrong. Unjustified corporate and government interference in network neutrality are equally big issues, and network neutrality fixes one of the problems and makes no difference to the other one.
Do you think Comcast should be prohibited from offering asymmetric broadband connections?
Suppose Comcast offered the following packages:
- 6 megabits down, 2 megabits up for $50/month
- 10 megabits down, 4 megabits up for $80/month
- 10 megabigs down, 4 megabits up, 20 megabits to youtube.com for $35/month
You're telling me that package 3 reduces consumer welfare and threatens the internet?
Suppose another package were added:
- 2 megabits down, 2 megabits up, all requests matching bing.com?q= redirect to google.com?q= ... price FREE.
How is consumer welfare being harmed? Any consumer who would choose this package values free internet more than the "right" to use Bing.
Or, suppose there is a mainstream movie about bittorent and suddenly Comcast decides to offer the following:
20 megabits down, 20 megabits up, no torrent traffic: $30/month. (with bittorrent it costs $80/month)
I'd argue that protocols have evolved to exploit network neutrality. Why should your ISP pay for Skype's supernode? Why should bandwidth hogging protocol users hog all the bandwidth?
Residential bandwidth sales are speculative, since the ISP doesn't have 20 megabits for every customer who is paying for it. The system works b/c people don't max out their connection and rarely use peak bandwidth at exactly the same time.
I think under this usage pattern, it's reasonable for prohibitions to exist. I also think that there should be higher tier bandwidth (like if you buy your own OC3) where you are buying all the bandwidth whether you use it or not. For such services non-neutrality makes no sense whatsoever.
The problem isn't so much that Comcast et al. want to offer packages 3, 4, and 5, it's that they want to discontinue packages 1 and 2 and replace them with crippled packages at the same price, with premium charges for YouTube, etc.
"Why should bandwidth hogging protocol users hog all the bandwidth?" Why don't ISPs simply charge by the megabyte, so that by the time they have too many bandwidth hogs on their network, they're bringing in enough money to afford network upgrades? "Bandwidth" isn't some fixed resource. If there's demand for more, the suppliers can provide more by either laying extra cables or upgrading the equipment at the ends of their fibers, and this kind of reaction can happen on the timescale of months, which makes it pretty fluid given that consumers also pay by the month.
True, but it's more a case of a small number of users resulting in tremendous loss of profits.
Suppose you open a restaurant and offer free bread and some customers come in and buy the cheapest item on the menu and eat $10 worth of bread. It's just not sustainable.
Which is why no restaurant claims to offer unlimited free bread.
There is no "but". If it were actually true that a small number of users consume so much bandwidth that it causes "tremendous loss of profits", then ISPs should drop the "unlimited bandwidth" claim.
But they want to keep that because it's just so nice for marketing.
What they really want is to claim "unlimited bandwidth" without actually offering it.
The problem is that Comcast wants to sell "20 megabits down, 20 megabits up, no torrent traffic" but market it as "20 megabits down, 20 megabits up". Which is actually a problem that doesn't require complete network neutrality to solve, but might benefit from government intervention.
Agreed that is a problem, but if you have noticed most of them have dropped the "unlimited!" from their slogans. I recall when my local cable provider was called Optimum Online Unlimited!. What they meant was that it was unlimited time, always on. Explictedly not unlimited bandwidth. I know this becuase I was cut off about 3 times in 6 months for using too much bandwidth. When you used to have to pay per minute, having an always on internet connection was "unlimited". Now everyone just assumes they pay 1 price per month for having access to the internet for the entire month. Would it make any sense to build the infrastructure so that every customer can saturate 20 megabits down every minute of every day of the month? Of course not. So they can't provide the model of always connected and unlimited bandwidth. It's one or the other.
I have an option of different speeds through my provider, and I'm OK with paying extra for the higher speed. Internet connection is more important to me than what channels I get on TV (which begs the question why I pay more for TV than internet). My parents would rather have more channels and slower internet. I don't see anything wrong with that.
In my comment, I said that asymmetric connections are not a network neutrality issue - by which I mean that proposed network neutrality laws don't / shouldn't cover them, so the article is bringing up a strawman issue.
Options 4 and 5 are network neutrality issues. If they are allowed, it is likely that they could be more cost effective for ISPs, and become the only options. Competition doesn't work for things like this, because most consumers don't understand the benefits of network neutrality.
However, deviations from network neutrality are bad for everyone in the long run, because it means that ISPs become gatekeepers, and slow innovation on the Internet.
If I were to use Skype, I would pay my ISP for traffic between my computer and Skype's supernode, and the owner of the Skype supernode pays their ISP for the traffic as well. The only time traffic isn't paid for is generally when it gets to the tier 1 networks, who don't charge each other because the exchange of traffic is mutually beneficial.
Given a fixed level of network hardware, bandwidth is a scarce resource, and economics provides a very good way to manage scarce resources - charge for them. This is exactly what happens at present. There is no need for any kind of ISP control over what sites are accessed or what protocols are used - and if too many people are using the bandwidth, ISPs can simply put up the price.
ISPs undoubtedly oversell bandwidth - but that is not an insurmountable problem for ISPs if there are network neutrality laws - they can either oversell bandwidth by less, or just stop overselling.
As one of the other comments pointed out, ISPs could just charge for bandwidth.
The current (speculative) pricing schemes and hypothetical non-neutral pricing schemes are simply alternatives to pricing per megabyte.
Why? Because ISPs believe that their product appears to have moe value when it's sold as x megabits per second for y dollars rather than x megabytes for y dollars.
Since most customers can't estimate how many megabytes they use each month, it makes little sense to price the product that way. Whereas many customers know that if they currently have 10 Megabits, if they upgraded to 20 megabits their downloads would be faster.
Comcast's boost pricing is evidence of the way ISPs are evolving to meet consumer demands. They give you lots of bandwidth for just long enough to let you download a song or a movie, but don't constantly make it available to you. This creates the perception of faster speeds for most customers.
For such services non-neutrality makes no sense whatsoever.
Basically the whole point of purchasing internet access is that you can use the sites and services you choose, just as the whole point of purchasing telephone service is that you can call the people you choose and receive calls from others.
Thus I have a hard time seeing how non-neutrality makes much sense on any sort of plan. If a customer is using the service to excess, charge the customer for use -- that's perfectly sensible. But saying "this customers used the service to excess, therefore we will disable part of the service for all customers" is nonsensical.
I think the key point is that customers are very different and that currently "broadband" is sold as a single type of product.
The person who downloads a few songs and videos and checks email wants a very different product from the person who is constantly uploading torrents.
Net neutrality is mostly just FUD though b/c with the exception of protocols that specifically exploit "shared bandwidth" pricing, it's unlikely to impact anyone (since consumers do like choice). Even my google/bing example is quite farfetched, b/c it's not likely worth enough to Google to subsidize bandwidth just to get search clicks, and in any case Comcast could sell Google analytics data without having to restrict anyone's choices.
I don't agree with the Cato article either, but this just sounds hopelessly naive:
"There is an obvious and simple solution to [players with a special interest to try to game the system] - transparency and openness in decision making."
Couldn't your argument be applied anytime someone objected to the unnecessary intrusion of government regulation? "Oh, we'll just have the government make decisions in an transparent and open way. Problem solved."
Because "the government" is an abstraction. In reality the decisions have to be made by specific people with specific interests and biases.
People with regulatory decision-making power are, by definition, powerful. So they have the interests of powerful people, which are unlikely to match the interests of most of the rest of us.
The foremost motive for an organization (like any organiam) is self-preservation. Transparency is usually perceived as a vulnerability that might allow enemies to attack.
to paraphrase "Internet access is not a right, so we shouldn't protect it". The reality is that so much human communication, including participation in democratic processes and a huge part of the economy happens online, and so it is reasonable to consider it a human right. The fact that it is relatively new doesn't change that.
On the one hand internet access and participation in digital communities is almost a modern form of free assembly. It's important for just the same reasons that free assembly is --- and threatening to power structures for the same reasons too.
The problem is that there are no "public spaces" in which to do this assembly --- all of the spaces on the internet are privately controlled. But this "modern free assembly" that's emerged is tremendously important, and we need to find some way to guarantee it.
6 - To paraphrase "ISPs should be free to put whatever they want in their contracts". The reality is that 'freedom of contract' is not fair when there is a huge difference in bargaining power, the markets can't sort it out for the reasons discussed above, and government intervention is appropriate.
Even if there actually is a huge difference in bargaining power, it does not follow that government intervention is appropriate. Existence of some sort of "unfairness" in the market does not mean (1) that restricting people from entering such contracts is moral and (2) that government intervention will result in a better outcome.
7 - To paraphrase, "Network operators own their equipment and should be able to do what they like with it". That is like saying gun owners should be able to do what they like with their gun, even shoot people. [...]
Not it is not. Shooting (innocent) people infringes on their rights. "Exploiting a market position", even if one believes it to be a reasonable issue, does not. The analogy does not apply at all.
The point here is that network operators to not have the obligation to use their own property to provide services to anyone, in the first place. There's nothing wrong if they decide to start offering those services for a ridiculously high price -- it's much better than no service at all. And there's nothing wrong if they also decide to offer bandwidth discounts to certain web sites.
> Favouring downloads over uploads for technical reasons - that is not really within the scope of network neutrality.
Why not? It blatantly discriminates against uploading services to the direct cash benefit of downloading services.
And that is the heart of the argument against network "neutrality": discrimination laws are always and everywhere a full-employment act for lawyers and bureaucrats. And AT&T will buy those lawyers and bureaucrats and use them to shut down competition.
> The market can't sort it out, because there are high barriers to entry, it is a finite market, and so the market is not infinitely efficient (and in some areas approaches a monopoly), and in practice there is little room for consumer choice influencing Internet markets except with regards to price, advertising, and to some extent reliability.
What market? The Internet industry is still in the wildcat stage, with most of the physical plant becoming obsolete and being discarded on a ~15 year schedule. They are trying to rigidly control a market that we all hope will be burned to the ground in ten years because the technology just isn't good enough.
> The reality is that 'freedom of contract' is not fair when there is a huge difference in bargaining power, the markets can't sort it out for the reasons discussed above, and government intervention is appropriate.
Indeed, which is why the government needs to keep competitive markets open and let them sort out the winners and losers. At this point, nobody knows what services are useful or valuable, so mandating some sort of ideological bandwidth management is pointless.
For example, Amazon might start renting space for a mini hard drive cluster in the telco neighborhood boxes, so that we get TV on demand from speedy local data farms. As surely as the sun rises in the east, Akamai would promptly get a restraining order preventing such a barbaric infringement of network "neutrality".
Of course their arguments don't stack up. They're professional conservatives. They are paid very well to systematically lie and distort for very specific purposes.
Uh, actually, they espouse GOP positions for money. They state that they are libertarians, yes, we know that. That doesn't mean that they don't support a lot of the exact same things that the GOP establishment does.
Going through it point by point:
* Their first argument against is that it is hard to define network discrimination. But it really isn't, and I can define it in one sentence: routing network packets differently depending on anything other than header information specifically designed for routers to process (IP headers in the case of Internet Protocol), or routing outgoing traffic based on destination addresses with the intention of deliberately slowing traffic to some destinations but not others.
* 2 - to paraphrase "Internet access is not a right, so we shouldn't protect it". The reality is that so much human communication, including participation in democratic processes and a huge part of the economy happens online, and so it is reasonable to consider it a human right. The fact that it is relatively new doesn't change that.
* 3 - To paraphrase "Network discrimination can be good, because it lets operators:
Stop immoral practices like spamming or 'hacking'" (presumably they mean cracking). That some people might do bad things not a good reason to censor the Internet. Favouring downloads over uploads for technical reasons - that is not really within the scope of network neutrality.
Conserve bandwidth - they give examples about usage caps, which are content neutral so outside of the scope of network neutrality. Specifically trying to slow certain applications, relative to another, is not necessary to limit bandwidth usage in a neutral way.
Promote brands - this is exactly why network neutrality is needed - network operators shouldn't be able to exploit their market position as gatekeeper to the Internet to favour one company over another.
Preventing 'theft' of service - except that if I pay for bandwidth, it is not theft to use it however I like, including allowing others to use it. Another good argument for network neutrality, not against.
Implementing a variety of network access charges to help to recover costs - ISPs can make money charging users for bandwidth in a non-discriminatory fashion, so there is no need for them to exploit their market position to charge people who aren't even connected to the network as well - and stopping them from doing that is a legitimate aim of network neutrality.
* 4 - to paraphrase "consumers might want more than a dumb pipe from their ISP, and if they go too far, the market will take care of it"; ISPs can offer services which are optional to consumers - that isn't a network neutrality issue, it only is when they become compulsory (e.g. you must use our nameservers) that it is. No consumer wants to be forced - at best they will tolerate it or it will make no difference.
* 5 - to paraphrase, "the market will sort it out". The market can't sort it out, because there are high barriers to entry, it is a finite market, and so the market is not infinitely efficient (and in some areas approaches a monopoly), and in practice there is little room for consumer choice influencing Internet markets except with regards to price, advertising, and to some extent reliability. While consumers benefit from network neutrality, very few are in a position to take this into account when choosing services.
* 6 - To paraphrase "ISPs should be free to put whatever they want in their contracts". The reality is that 'freedom of contract' is not fair when there is a huge difference in bargaining power, the markets can't sort it out for the reasons discussed above, and government intervention is appropriate.
* 7 - To paraphrase, "Network operators own their equipment and should be able to do what they like with it". That is like saying gun owners should be able to do what they like with their gun, even shoot people. It is reasonable to put in place laws against certain types of unreasonable exploitation of a market position, and property rights don't extend to using the property to break those reasonable laws. The argument here seems particularly disingenuous.
* 8 - To paraphrase, "network neutrality will discourage innovation and investment". The reality is that network operators can still profit from charging users for access - and that is enough to encourage investment. The main limit on ISPs profits is likely to be competition, and all ISPs are subject to the same laws - so no service that was profitable without network neutrality is likely to become unprofitable because of network neutrality.
* 9 - to paraphrase "introducing laws around the Internet will encourage players with a special interest to try to game the system". There is an obvious and simple solution to this one - transparency and openness in decision making. Gaming works both ways - gaming to not introduce network neutrality because it is in corporate interests anyway.
* 10 - to paraphrase - "putting in place Network Neutrality laws will start down a slippery slope towards government control of the network". Nothing specific to any network neutrality law makes it easier to change laws around the Internet later, so the argument is wrong. Unjustified corporate and government interference in network neutrality are equally big issues, and network neutrality fixes one of the problems and makes no difference to the other one.
All in all, none of their arguments stack up.