Net Neutrality vs Freedom

published

2014-05-15

In my last post, I discussed some problems with the general concept of net neutrality, but in this post I want to discuss net neutrality more concretely and in the sense that most people are talking about online.


The general assumption is that every user of the internet, whether they are a content provider or a consumer, pays for their connection to the internet at a certain service tier, and that covers all their network costs. This is a nice simple model. For simplicity, we are ignoring the costs associated with the transmission through the internet, which is bundled into the costs each user pays.

A packet traveling between X and A will be prioritized according to X’s service tier between X’s server and the internet, and according to A’s service tier between the internet and A’s house.

Net neutrality suggests that because the whole link has been paid for by one party or another, it doesn’t make sense to “double dip” by charging both X and A for service over the same part of the network, namely the ISP’s part of the network.

Now in a sense, both X and A are users of the ISP’s network. Every packet on the internet has a source and a destination, and both the source and destination benefit from the connection. Similarly, cell phone companies in the US charge by the minute for both incoming and outgoing calls. This could be considered double-dipping, but since the situation is already so symmetrical (both parties talk and listen), it actually seems more fair than just charging the caller as some other countries do.

So even though it may be A who is watching a movie on Netflix (server X), Netflix is simultaneously leveraging the ISP’s infrastructure to run their business. But customer A already paid for internet access of a certain tier, so a net neutrality advocate would suggest that customer A has already effectively paid on Netflix’s behalf.

The question here is “Has customer A actually paid the total fee for using the ISP’s network on behalf of X, Y, and all other remote sites?”. If A’s contract with the ISP states that the service is “remote party neutral”, then they have paid for it and it would be a contract violation to filter based on the remote party. In this case, there is no need for net neutrality regulations because standard contract law already guarantees that the ISP can’t prioritize connections to X over connections to Y. However, if the contract does not stipulate this, then it may be the case that A hasn’t actually paid for such a privilege. The ISP may have offered a lower rate subject to certain restrictions.

Now why would the ISP even bother to offer a restricted plan? One possibility is that they are also a cable TV company and they want to block Netflix because Netflix is costing them cable subscribers. I know a lot of people get upset by this, but if you think about it, that should be their right. If you do the work of laying down cables and launching an ISP, you should be able to do whatever you want with it, even run your business into the ground with poor service. Sure, it sounds like it will be bad for consumers, but sometimes it’s good to give a company enough rope to hang itself because that will make it easier for competitors to enter the market. If you force a company to be just good enough to deter competitors from entering the market, you may actually be helping a bad company retain their monopoly.

But realistically, I don’t think even bad ISP companies are that dumb. The reason why they are interested in charging content providers is because their current business model is economically unstable. By unstable, I mean that there are alternative business strategies that would outcompete them if implemented. This concept is the business analogue of the “evolutionarily stable strategy (ESS)”. It is unstable because some customers are getting a better deal than others. People are paying for download rates, but they are not paying based on the quantity that they download. So those customers who download a larger total quantity of data are getting more service for the same price as other users on the same tier. If a competitor ISP charges people for what they actually use, the lower utilizing customers will switch to the competitor for the lower prices and the original ISP will have to raise its prices for the higher utilizing customers because the lower utilizing customers are no longer subsidizing them. So in this case the new ISP gets all the lower utilizing customers and half the higher utilizing customers (since both ISPs end up with the same price for them). That means that the new ISP gets more customers and outcompetes the old ISP.

Now probably the simplest option is for the ISP to just add a per-GB fee to their pricing structure. That would be an economically stable strategy from a pricing standpoint. But maybe the ISP assumes that people just don’t like per-GB fees, perhaps due to their market research. So then the ISP notices that their users fall fairly neatly into two categories: those who use streaming video services and those who don’t. The ISP realizes that they can effectively charge per GB without adding per-GB fees by charging a flat rate that is higher for customers who use streaming video. This strategy mimics the economically stable strategy without upsetting customers over per-GB fees. It is a business optimization, created in response to market pressures.

What about startups? If content providers have to pay every ISP in the country for fast-lane access to users, won’t that make it expensive to start a new streaming video service? Well, it already is quite expensive to start bandwidth-intensive internet companies. There are always barriers to entry in a business; that’s no reason to make laws that arbitrarily threaten other businesses. A startup might have to rollout their services one ISP at a time if they don’t have enough funding to cut deals with all ISPs at once. And really that is the worst case scenario. More likely the ISP will only be throttling the big names like Amazon and Netflix, which would actually provide a competitive advantage to startups because they would get fast-lane access without paying the fees that the big names pay.

And we haven’t even considered the unintended consequences of net neutrality regulations yet. For example, if a group of hackers launched a cyber-attack on one of the ISP’s customers, flooding the ISP’s pipes with packets and bringing the ISP’s services to a crawl, are we saying that the ISP should be legally forbidden from doing anything to fix the situation? That’s what straight-up net neutrality would suggest. Even if you wrote an exception for cyber-attacks into the legislation, what happens when a legitimate service comes out that effectively does the same thing? Perhaps a service that pings every computer on the internet many times per second, like a search engine spider, except that it indexes all nodes on the internet instead of just web servers. The customer isn’t requesting that traffic at all, so does it still make sense to say that this traffic is already paid for by the customer, or is it a form of freeloading by the remote service? You can probably work around this issue by making a net neutrality exception for anything that the customer didn’t request, but now it’s getting more complicated because you have to define what that means and it goes to show that making laws can cause problems you didn’t anticipate.

The bottom line is that businesses should be free to do business and to try any optimization they want as long as they aren’t committing fraud or violence. If they provide terrible service, the free market will fix the problem. Monopolies almost never last long and they only exist at all if they are doing a decent job and/or the government is helping them. And yes, governments are helping monopolies by making permitting processes difficult and rights-of-way access expensive. Piling on more government legislation (in the form of net neutrality regulations) is not the solution to having too much government regulation. The solution is to fix the problem at its core, reduce regulatory barriers, and encourage competition. I’m confident that we’ll retain remote-neutral internet access if that’s what the people really want. Promote freedom, not net neutrality.

Icons in the image are from the Oxygen Icon Pack, except the server which is from Itzik Gur.