Why 'zero rating' is the new battleground in net neutrality debate
cbc.ca
cbc.ca
Just like the US, but so much worse.
All in all, I think the CRTC has taken a pretty good approach. Perhaps the third-party only rule should be extended to refusing exclusivity arrangements, but a complete ban is over the top in my opinion - it discourages passing on savings made by peering with companies to consumers.
You (and startups) are basically asking for a subsidy that doesn't exist for anyone else. No one is giving Jolla a hand in taking on Samsung or Apple in the phone business.
If they do (by incentivizing people to conserve bandwidth on, say, mobile connections) then they need to be consistently enforced, with no exceptions.
If they don't serve a purpose for consumers then they should be abolished anyway.
If Spotify and WIMP use equal amounts of data then they should cost you, the consumer, the same in transfer costs. Anything else just ends up with all consumers subsidizing the incumbents no matter what vendors they choose (like the often criticized "Microsoft tax"), unless they only go with startups that do not have these agreements in place (yet).
> You (and startups) are basically asking for a subsidy that doesn't exist for anyone else. No one is giving Jolla a hand in taking on Samsung or Apple in the phone business.
Oh yeah, let's never try to improve things in new markets that work differently.
You can choose to imagine cellphone hardware as a utility but clearly it is not.
So they shouldn't be allowed to pass the savings on to the customers who actually use the service in question and thus benefit from the colocated server? And this is supposed to be more "fair"? And somehow the ISP's rate model for their customers is yours to dictate?
Consider if ISPs actually charged customers by data transfer, the way many large datacenters do. What actually costs the ISP money is data transferred outside the ISP's network, going over peering connections. Many datacenters don't charge for bandwidth within their own network; for instance, bandwidth between AWS nodes in the same area is free. So why can't an ISP use the same model, and only charge for bandwidth going over their peering connections but not for bandwidth that stays inside their own network (such as to the YouTube or Wikipedia mirror they're hosting)?
For the record, I do believe in network neutrality in the sense that ISPs should not be artificially limiting bandwidth to particular Internet services (though I'd prefer to see that solved more naturally through ISP competition, but there's a severe lack of local ISP competition available). I don't, however, see anything wrong with setting up additional peering/colocation/etc arrangements that are designed to be mutually beneficial to the ISP, their customers, and the services whose data they're carrying.
Zero rating is not about passing on cost savings, it's about market power and discrimination.
Also, datacenters do not usually charge by data transfer. Just because Amazon does it (and overcharges by two orders of magnitude) does not mean others do it.
The article points out the big downside: Zero rating takes away a strong incentive to raise bandwidth caps, as low bandwidth caps makes it more attractive for content providers to negotiate zero-rating deals.
The net result is a market heavily distorted in favour of a small set of major content providers.
> it discourages passing on savings made by peering with companies to consumers.
Cellphone data costs rarely have much relation to transit cost. Ex. my provider in the UK offers 2GB at 20 pounds a month. The actual data cost from a really expensive transit provider in a non-carrier neutral data centre for 2GB is about 7p. 0.07 pounds.... With some peering, decent volume, and sourcing bandwidth in a carrier neutral location where there's decent competition would bring that down to a tiny fraction.
Australia is a country which imports the majority of its content (submarine cables are not cheap), so smaller ISPs have an incentive to direct eyeballs to 'friendly' content sources - those who peer at major peering exchanges locally, or directly.
At least with the independent providers, this has been used for good - unmetered Linux mirrors, Steam, Netflix - as these providers do not have any content of their own, unlike Telstra (who holds 50% of Foxtel, alongside Newscorp).
[1] this 'cartel' was created under pressure of the competition regulator/ACCC in the late '90s, ironically. The criteria isn't based on any merit, just the largest market share holders at the time.
The other problem was just terrible speeds; best I could get in a central district was ADSL2+, with about 16/0.8 mbit d/u, and that's not including the terrible latency issue you had to deal with for a whole host of US and EU based websites and services.
Then again, I'm back in London now with a 150/15 mbit d/u connection for under £30, no caps in sight.
They shift price around in a way that creates a local maxima or suboptimal equilibrium.
> "Zero rating isn't great for consumers, as it has the potential to distort consumer choice in favour of choices selected by an ISP," Netflix spokesman Cliff Edwards says. "We'll push back against such efforts, but we won't put our service or our members at a disadvantage."
I think you just don't see the problems it creates.
> It has largely been consumer positive, allowing premium tier ISPs to differentiate themselves by making the effort to peer with service providers, and consumers benefit from the free quota.
They already have the incentive of the fact if they don't peer with service providers, they won't actually be able to connect you to the rest of the internet at a reasonable speed. Australia's market for the internet is worse than the US because of things like this distorting it + the distance from the rest of the world.
I would never setup a server in Australia because of the ISPs. I would stand them up on the west coast of the US instead precisely because of their behavior. I think you don't really understand how much the ISPs of Australia have made the entire Australian market completely uncompetitive with the rest of the world.
Here is a random example of a company that operates in Australia and abroad.
https://www.vultr.com/pricing/
> Bandwidth quotas listed are for North American and European locations. Overage is priced at $0.02/GB in North America, $0.02/GB in Europe, $0.05/GB in Tokyo and $0.10/GB in Australia. Full plan details are listed on the customer portal.
Why do you think Australia costs $.10/GB while the rest of the world [outside of Tokyo] costs $.02/GB?
> All in all, I think the CRTC has taken a pretty good approach. Perhaps the third-party only rule should be extended to refusing exclusivity arrangements, but a complete ban is over the top in my opinion - it discourages passing on savings made by peering with companies to consumers.
It has the opposite effect. It makes things more expensive. The only reason it appears "cheaper" in the Australian market is because they decided to screw not-consumers in the Australian market even harder.
You mention sports, but that's a total anomaly, because upstart sports franchises are not common and they tend to be big entrenched players anyhow, due to the massive physical infrastructure required to run a league.
Sporting events is just not a good example of why zero-rating should be any sort of norm. And I can't think of a good one.
It only feel like a good thing because Australia has an abnormal telecom market which includes such strangeness as quotas even on fixed broadband.
Without quotas zero rating is a non-issue.
There are over 800 telcos in the world. Even if zero rating was free to the app/service provider, just negotiating with each and every telco would be a huge burden.
Zero rating is, however, neither free nor available to all app/service providers. Thus zero rating creates both a toll booth and a gatekeeper who gets to discriminate against app/service providers and pick winners and loosers.
what if 3rd parties where allowed to buy zero-rated URLs or IPs in "bulk" from all the carriers in a geography, making it much easier to contract all mobile carriers?
preventing companies to provide zero-rated services increases the digital divide as it inhibit people who cannot afford to buy a data plan to use services. In many areas of the world, the majority of mobile consumers are on pre-paid plans and have zero credits most of the time (most countries have calling party pays, so a smartphone with zero credit is still very valuable to receive calls)
While nominally better, it would just be like putting lipstick on a pig. There would also be serious real world problems with actually enforing such a regimen.
> what if 3rd parties where allowed to buy zero-rated URLs or IPs in "bulk" from all the carriers in a geography, making it much easier to contract all mobile carriers?
My answer is pretty much the same as above, only this one is a bit more ambitious. It's basically like suggesting world peace as a solution to global conflicts. Getting all the carriers in a geographic region on board, implementing, rolling out and enforcing such a thing would almost be an effort worth a Nobel price in itself.
> preventing companies to provide zero-rated services increases the digital divide as it inhibit people who cannot afford to buy a data plan to use services. In many areas of the world, the majority of mobile consumers are on pre-paid plans and have zero credits most of the time (most countries have calling party pays, so a smartphone with zero credit is still very valuable to receive calls)
Offering zero rating as a solution to this problem is a false choice. Zero rating is providing a means of communication for free AND limiting what it can be used for. The correct choice is to provide the same means of communications for free and NOT limiting what it can be used for. That is how you remove the digital divide.
If need be this free mode of communication can be limited by speed or by amount, as is customary in the mobile world. What is not reasonable is to limit what can be done with it.
TL;DR zero rating is evil, give users a free tier or quota instead.