Sun 21 May 2006
Net Neutrality is a deep issue
Posted by David under Uncategorized
[4] Comments
Do you remember when I said “[Everyone is right about
Alito](http://davidernst.net/blog/2006/01/24/everyone-is-right-about-alito/)”?
Well, everyone is wrong about Net Neutrality. Or, put more fairly,
everyone is right about some of the things that they say, and wrong
about other things that they say. Very little of the hailstorm of
opinion circulating on the subject is free of spin, overstatement or
falsehood. Almost all of the opinion is representing either a vested
interest or an ideological interest. I work for a small ISP, which
perhaps puts me in an interesting position: I understand (imho) pretty
well what’s going on, but the degree to which my interest in the
matter is vested is pretty minor. Furthermore, the question of the
government’s role here is deep. So, I offer my reading of the
situation, in hopes that it will help some folks to develop an
opinion.
First, the issue is pretty difficult to grasp when you explain it
abstractly. Search Google News for “net neutrality” and you’ll find
countless attempts, most of which are difficult to follow and already
include some of the opinion that motivated the writing of whatever you
are reading. I’ll sum it up simply by saying that some congresspeople
are working to pass a law that would require companies who sell
Internet service to provide equal access to any Internet information
provider. So, your ISP couldn’t offer preferential treatment to
certain web sites or Internet services over others. Sounds reasonable
eh? But… Preferential treatment? What would that look like? Most
people (at least, Net Neutrality advocates presenting their case to
the mainsteam media) describe it as “loading slower”. Like, if Google
doesn’t pay an ISP, then http://www.google.com will load more slowly
on your computer. To me, this doesn’t give a very accurate impression
of what’s really at stake here.
So, rather than the abstract description, consider this historical
description: From the beginning of the Internet through right now, if
a packet of data is addressed to your computer, ISPs have never taken
into account where that packet came from, or what purpose it is for,
when they send it to you. They have their routers configured to send
it to you at the speed you signed up for, and when everything is
working, that’s what they do. The only exception to this that I can
think of is content filtering. Perhaps you’ve signed up for an
Internet service that blocks pornography. This kind of filtering is
technically very different, but more important than the technical
difference is the fact that you, the one *paying* for the service,
*asked* for that content to be filtered.
So, for the first time ever, some Internet Service Providers — who,
nowadays, are mostly largest Cable TV and Telephone companies — are
making noises about taking the source of a data packet into account
when they decide if or how it will be sent to you. In particular,
they are talking about charging content providers for some assurance
of how your requests for information will be delivered to you. The
most telling example of this that I’ve seen was in [this Business Week
interview with AT&T (then SBC) CEO Edward
Whitacre](http://www.businessweek.com/@@n34h*IUQu7KtOwgA/magazine/content/05_45/b3958092.htm):
Business Week: How concerned are you about Internet upstarts like
Google, MSN [an Internet upstart?? -dre], Vonage, and others?
Whitacre: How do you think they’re going to get to customers? Through
a broadband pipe. Cable companies have them. We have them. Now what
they would like to do is use my pipes free, but I ain’t going to let
them do that because we have spent this capital and we have to have a
return on it. So there’s going to have to be some mechanism for these
people who use these pipes to pay for the portion they’re using. Why
should they be allowed to use my pipes?The Internet can’t be free in that sense, because we and the cable
companies have made an investment and for a Google or Yahoo!
or Vonage or anybody to expect to use these pipes [for] free is nuts!
Personally, I think Mr. Whitacre would have been wise to have had his
spin-doctors in the room during thie Interview, because it’s this kind
of reckless talk that really sparked the debate. “use my pipes free”?
Wait a minute. If I buy Internet service from you, that pipe you just
called “yours” is the one that I think of as “mine”. You think it’s
yours because you (which really means the corporation that employs
you) paid for it and built it. I think it’s mine because I’m paying
you a fee every month to get information from the Internet with it.
So, it ends up being a lot like a landlord/tenant relationship. The
landlord is reasonable to try to protect the value of their
investment, the tenant is reasonable in thinking that they deserve
freedom in how they use the thing they are renting.
Now, my guess (and I want to emphasize that this is a guess) is that
this has been a key element of AT&T’s Internet business plan for a
while. For a long while (most of it when they were known as SBC),
AT&T has been offering DSL service to end users at prices that were
unbelievably low. As someone who works in the field, I couldn’t even
believe them. In fact, I’ve personally never believed that they were
making money on their DSL offerings. I felt like their low prices
were motivated by a long term strategy. They were behind in the
broadband race, the cable companies had way more market share. They
had to play catch up. But, they have money, so why not sell at a loss
for a while to get some market share? Besides, they’ve got to be
petrified of the idea that people could have cable modem service and a
cell phone and then just get rid of their landline (as many people
have). DSL from SBC/AT&T has thus far always meant that you still
have an SBC/AT&T phone line. When people stop paying AT&T monthly
fees, that giant could fall unbelievably quickly.
Anyway, all of that is stuff I’ve believed for a long while. What I
didn’t see coming was what Whitacre describes above. Google has lots
of cash right now. So does Microsoft. Maybe Yahoo does too. Why not
charge them! If AT&T gets enough customers, that customer base starts
to look valuable. Google makes almost all of its money showing ads to
end users, so the companies that connect those end-users to the
Internet have a bargaining chip. Indeed, it could be a perfectly
reasonable model of how the world could work: End users pay a
ridiculously low price for really fast Internet access, and ISPs make
up the difference then some by getting some kickback from the people
who are making money by being able to connect with the end users. To
be honest, it’s not that different from the original NetZero model of
paying for the Internet Service by showing the end user ads.
The problem that I see with this is that it violates my understanding
of what I’m buying when I buy “Internet service”. Yahoo! provides
information via the Internet, right? And I am buying Internet
service, right? So, shouldn’t I be guaranteed that Yahoo! information
that I request will come to my computer without having to pass any
sort of judgement of my ISP? Shouldn’t Yahoo! — or, more
importantly, the web site of the small Internet startup company
without billions of dollars in the bank — know that it can reach
anyone who has “Internet service” without having to pay that person’s
service provider?
So, there’s a decent argument for the Net Neutrality bill. But,
before we get too far into this mindset, there’s another key factor:
IMHO, it’s highly unlikely that any ISP would distort how things work
right now. All this talk of “Google loading more slowly” doesn’t make
much sense to me. They aren’t going to add a five second pause to all
information coming from Google. Or anyone else. They’re risking
enough by talking the way they are, if people’s worst fears become
true, customers will bail as fast as they can. My reading is that the
debate is not about the Internet as we know it now. It’s about the TV
version of the Internet to come. So, a word about that.
Back when dialup modems were the most common thing, there was a lot
more concern about how big a website was, which mostly meant images.
Text can be downloaded so much faster than any human can read (even
over a modem) that you didn’t have to worry much about text. But
images made web pages load slowly over a modem. And we didn’t like
that. Those of us with broadband are pretty much over this. Even
image-heavy websites load comfortably fast, and I hear a lot less
complaining about the “world wide wait” than I used to. So, now
what’s uncomfortably slow (that lots of people want) is audio/video
media. You’ve got your choice of taking something that’s poor quality
compared to TV (streamed media via Real, Quicktime, etc) or something
that’s great quality but takes hours to download (media files
downloaded by something like BitTorrent, iTunes, etc). The main
benefit of the next quantum leap in Internet bandwidth is probably
going to be that the Internet is more like TV. Those of us who prefer
the Internet as it is now to TV as it is now have trouble getting
excited about this. But, the truth is that hypertext web pages as we
know them are not going to go away, they’ll work just fine on a faster
Internet. But so will video.
This, in my analysis, is what the big ISPs care about. What they
really want to be able to do is say “Ok, TV-like content takes up huge
amounts of bandwidth, but people want it. So, we want to upgrade our
customers to 100Mbps by installing fiberoptic lines to their homes,
but then we want to sell them on-demand HDTV-quality video over that
infrastructure, and we don’t want to have to compete with outsiders
who want to do the same thing, because we’re the ones paying for all
of that fiber. And besides, we don’t want to have to pay billions of
dollars on the other side of our network so that third parties can get
through to you at those kinds of speeds.” If Mr. Whitacre had phrased
his position this way, there might not even be a bill before congress
right now.
So, what to do? Well, to me, the debate really is a deep question
about free markets. Personally, I don’t mind if AT&T starts selling
some Internet Service that’s fast enough for great AV content from
select providers, but that keeps others out of selling that kind of
content. However, I don’t want to pay for Internet service from an
Internet Service Provider and then find out that the funny video that
my friends watched “on the Internet” won’t play on my machine because
my “Internet Service Provider” doesn’t allow that kind of content from
that provider.
So, I believe where my own opinion is headed with this is towards one
of my favorite types of government regulation: labelling. I love that
most of the food that I buy has basic nutrition information on it.
It’s easy and cheap for businesses to comply with, has a huge public
benefit, and doesn’t disturb the market economy. If it was just
agreed upon (by government decree or not) that “Internet Service”
meant the type of source/purpose-blind routing to the end user that
we’re used to, but that other services could also be offered, that
seems at first blush like it would solve the problem. AT&T and their
ilk could market “1.5Mbps Internet Service, and 100Mbps AT&V Video
Service, all on one line for one great price!” I don’t think anyone
would complain. Consumers would know that they weren’t going to be
able to use services from third parties that required more than 1.5
Mbps, and if they wanted more than that, they could shop around. If
outside companies wanted to offer services that required huge (by
present-day standards) amounts of bandwidth, they could consider
paying AT&T to be able to reach their customers.
What no one wants is for a consumer to buy Internet Service and then
find that they can’t get what they understood to be part of “Internet
Service”. The funny thing is that the Big ISPs don’t want that
either. It’s completely wrong when savetheinternet.org says “Right
now AT&T and others want to take away your choices and control what
you can do and watch online.” AT&T and those others just want to make
money, they don’t care if I have choices or not. Not caring is a lot
different than “wanting to take away choices.” But it’s equally wrong
when dontregulate.org (I can’t help but point out the incongruity of
such a forceful a domain name with a tagline of “make up your own
mind”) says that the Net Neutrality supported “want the government to
take control of the Internet”. No, none of them want the Government
involved, they just don’t want a oligopoly to take control of the
Internet, and the government is the most obvious way to keep that from
happening.
The free marketeers will argue that the market always works itself
out. I like free markets in general, but I don’t think they always
work themselves out. Here’s how they might in this case. Imagine
this fight: AT&T and Comcast start doing what people are saying they
will do, and slow down content from, say, video.google.com, to the
point where those videos won’t play on the computers of their
customers. Google says “Oh yeah? well, we think your customers care
about reaching us, so we’ve just decided that we’re going going to let
AT&T customers get to ANY of our content.” Now Comcast customers see
a page that says “Sorry, Comcast restricts content from Google, and as
such we are no longer providing service to comcast members. Here’s a
list of ISPs in your area that do not restrict content based on host.
For a limited time, Google is offering a $50 rebate to people who
switch from Comcast to one of these ISPs, to help defray any fees that
might be incurred in switching providers.” It may seem crazy, but a
couple of years ago, something similar happened between a cable
service provider and a major TV network stopped providing video to a
cable TV company (I’m sorry I don’t remember which companies were
involved in this. If anyone else does, please let me know). And, I
bet this would work. AT&T and Comcast, who are natural competitors,
would be in a prisoner’s dilemma situation, where the first one of
them to break the pact would benefit. Such things would make for
great entertainment, but in the meantime a lot of people would have a
lot of trouble getting satisfactory service out of the Internet.
Might it be worth a bit of government regulation so that we don’t have
to live through that?
Like I say, I currently think that the best approach would just to be
to give clear definition to what is expected of “Internet Service.” I
think those who oppose the Net Neutrality bill have a point when they
say that the legislation proposed thus far is short-sighted and
dangerous because the congress people who would be writing it and
voting for it don’t really understand the issues. However, the bill’s
supporters also have a point that a power grab from the major
telephone and cable companies (who have
annoyed us all in countless ways over the years) could really damage
what we understand the Internet to be. In short, I don’t trust the
government or the large ISPs. But I do think it’s reasonable to think
that if they don’t get their way, the ISPs will be less likely to
build out fiber-to-the-home, because that would be ridiculously
expensive and it won’t work on modern business plans. Of course, I’m
happy with the Internet as it is now, I don’t really care if it
becomes more like TV, so maybe I’m biased. :)
One final comment: much of the mess of modern US telecommunications
system stems from the 1918-1983 government enforced monopoly held by
AT&T. Since 1984, the “Baby Bells” have used all kinds of shady
techniques to obstruct competition from other providers. Their empire
was built on government “regulation” to a degree seen today probably
nowhere except possibly Amtrak. For them to take a “free market”
stance on any issue should make them blush.
4 Responses to “ Net Neutrality is a deep issue ”
Trackbacks & Pingbacks:
-
[…] All this is basically to serve as a reminder that techological advances are not always limited by technical issues, sometimes it’s our physiology that limits things. I’ve hinted previously that once we have enough network bandwidth to do high-quality video and audio in real time, there won’t be much incentive to go faster. I’m not ready to stick my neck out on that yet, but I will say that I don’t see much demand for anything that would require more bandwidth than that. Still, that’s quite a ways away… […]
Wow, as I’m sure you can imagine, I’ve also got a lot of thoughts about this.
It has been interesting watching the (lack of) development around a concept called Quality Of Service, of QoS. The first definition Google gives for QoS is as follows : “On the Internet and in other networks, Quality of Service (QoS) is the idea that transmission rates, error rates, and other characteristics can be measured, improved, and, to some extent, guaranteed in advance.”
Sounds like a good thing to be able to do! For example, if you are using a Voice Over IP (VoIP) service like Vontage, you don’t want your phone call breaking up because your partner has just started downloading video. And I think everyone agrees that 911-type calls should take priority over any other type of network traffic. Another example is a doctor controlling a robot that is performing an operation two-thousand miles away; that network traffic should get priority.
But QoS also enables the type of issues we’re starting to see develop politically as well. With QoS, the First Tier Internet Service Providers (the big boys that own infrastructure backbones) can also start pricing based on the quality provided. I agree with David that there is most assuredly nothing wrong with this in principal. But I would also say that like the telephone, gas, water, and sewage, the Internet is now a utility and as such, there should federal regulation mandating a minimum level of access quality. How that would get implemented is a hornet’s nest, but I think it is a requirement. Otherwise, I think instead of providing “premium†service above and beyond a standard performance level (which would cost significant cost to make premium service available) the Tier 1 and Tier 2 ISPs are going to be much more interested in providing a “crippled†basic service (extremely slow and/or not full access) and a “full†service which is very little different than what we have now. The ISPs will be able to generate more revenue with little infrastructure cost. It is always cheaper to cripple service rather than enhance it.
And I have major trouble with the ISP’s general argument about Google. “They are making money hand-over-fist, and we should get a cut of that!†Well, the ISPs do! Google does not get its network access for free from anyone. And the more traffic Google generates, the more Google must pay for larger “pipes†to provide that service. If the ISPs believe they are getting ripped off, then they can easily re-negotiate the service agreements with Google.
Google, by the way, has seen this coming. And they are doing something absolutely unthinkable in response; building out their own infrastructure. The scope of building an independent network infrastructure is unthinkable. Except evidentially, if you are Google.
But there is a more core issue that bothers me about all of this; and that is the difference between a distributed and centralized infrastructure.
The phone network is centralized. Your phone (and I mean no offense) is dumb. No matter how many buttons your phone has, there are only twelve that make a real difference, 0-9,# and *. Add to that “on hook/off hook†and that’s what you get. Anything else is pretty much just a shortcut to those features. (FLASH, for example, is just a very quick on hook/off hook.) All of the real intelligence to handle three-way calling, caller ID, call block, or any other service is handled by centralized systems at the phone company. All of the power is in the core.
The Internet is distributed. The “dumb†part is actually INSIDE the network. This means that the Internet in the core is historically just concerned with how to route packets from one place to the other. The Internet doesn’t care at all about what is being routed, it is just concerned with the best way to do it. Any “intelligence†such as websites, databases, file shares, or clients are on the “edge†of the Internet – as end-nodes. All of the real power is at the edge.
The real impact on power being on the edge verses the core is on innovation. The reason we were able to have Googles or Vontages at all is because absolutely anyone can come up with a cool idea, stuck it on the edge of the Internet somewhere, and if they can convince others to play along, launch a new service. The Internet doesn’t care – as long as the people with a good idea meet the open standards, the packets will flow. This also means however, that people who want to write viruses, worms, Trojans, and other nasties will always have ample opportunity to inject their badness into the Internet.
QoS is a beginning step to transforming the Internet from an edge-powered infrastructure to a centralized infrastructure. From my perspective, this isn’t necessarily a bad idea – but it is one with consequences. On the PRO side it will make for a more stable and controlled experience; you can trust more that the content you will receive will be “clean†of viruses, you will have guaranteed service for those 911 calls, and overall a more rich media experience if you have the money to pay for it. On the CON side, the innovation which has brought us the majority of services we now think of as making up the Internet will stop. YOU can’t write a program that will allow you to create a concept like three-way calling on the phones. The Phone Company needs to have that idea and program it themselves.
I don’t have any brilliant endings to this response. Honestly, I don’t think there is one. This is a very complex issue and there are no simple answers. What is maddening to me however, is that the logic of this discussion will actually have very little to do with the outcome.
Wow. Well, my main thoughts are these: 1.) Hold on to those ideas, so that you can tell me more about them over New Years. 2.) I’m really glad that I have friends like dre and mpburton to stay on top of these issues so that I can focus on things like the Israeli-Palestinian conflict and 3.) I’d better stop lazing around and keep up with the papers so that I can hold up my end of that bargain.
Also, aside from Alito, QoS, and other issues, I believe strongly in the serial comma. I just thought I would mention it.