Free Press is asking the FCC to consider a number of changes to the NPRM Net Neutrality regulations which they claim will “promote investment”. But once we examine their proposal in detail, we find that it will produce just the opposite and devastate the U.S. economy. Not only would Free Press preclude broadband providers from innovative new business models and economic opportunities, they would substantially undermine their existing business models and revenue streams. Yet despite all this, Free Press insists that their proposals would not deter broadband companies from investing money but that it would spur new dotcom investments at the edges of the network.
The Free Press proposal demands the following regulations
* Outlaw paid prioritization services
* Outlaw managed service prioritization and bandwidth allocation
Free Press argues that Net Neutrality rules that AT&T agreed to in 2006 didn’t negatively impact their investments on network infrastructure and therefore these current Net Neutrality proposals won’t have any negative impact either. While it’s true that the temporary FCC rules AT&T agreed to in 2006 did not reduce AT&T investments, it doesn’t mean that there was no adverse impact. Furthermore, those FCC rules in 2006 were not nearly as restrictive as the rules being proposed today so the comparison between then and now isn’t remotely accurate.
Outlawing paid prioritization services
Free Press insists that “Pay-for-Priority” services are bad with the following argument:
“Pay-for-Priority” — Under this scheme, third-party content and applications providers would compensate ISPs for prioritizing their traffic over all other traffic flowing across the ISP’s network. But this model is faced with an immovable barrier: the routing of Internet packets is a zero-sum-game; during times of congestion, prioritizing one packet deprioritizes all others. This practical reality firmly bounds the possibilities of the pay-for-priority business model. In practice, this means that in order for this model to work, congestion will have to be widespread. This in turn means that ISPs will likely form exclusive paid-priority relationships, resulting in the Balkanization of the Internet.”
Free Press seems to be stuck on the concept of the zero-sum-game. They wrongly assume that prioritization is harmful to non-prioritized applications and they were thoroughly debunked here on every key assertion. Engineers know that network prioritization is not a zero-sum-game and that a prioritized network is actually more fair and neutral than the “dumb pipe” First In First Out (FIFO) network that Free Press foolishly worships.
Furthermore, ISPs generally connect premium peering customers to new dedicated infrastructure without piling them onto the existing general purpose infrastructure. That means premium services do not decrease the available capacity to standard customers but just the opposite because premium customers are offloaded to new infrastructure. As for the shared last mile, ISPs spend billions of dollars a year ensuring that capacity grows with traffic load so that customer experience isn’t degraded because ISPs can’t afford to lose customers and reputation. The reality is that more premium services put more load on the broadband network which creates a need to upgrade the network, and the premium services provide the necessary revenue to fund the upgrades.
Free Press also argues that:
“Without Network Neutrality, ISPs will have a strong incentive to reduce investment and make congestion commonplace in order to extract revenues from content providers willing to pay to avoid traffic delays.”
The fallacy with this argument is that it assumes that ISPs can permit their network performance to degrade without alienating existing customers and driving them to a competitor. The reality is that bad service equals loss of customers which far outweighs the cost savings of skimping on infrastructure, and it endangers the long term reputation of the ISP. All ISPs monitor their network load levels and the respond to increasing traffic by adding capacity and they spend tens of billions of dollars every year upgrading their networks. If ISPs could actually get away with not upgrading the network in response to additional loads, they would surely refuse to spend billions of dollars on upgrades. It is an ISP’s fiduciary responsibility to their shareholders to avoid spending billions if there is no risk to their customer base or reputation.
The reality is that ISPs must deal with increased congestion and according to company reports and data from the CTIA and NCTA, ISPs spent $60.7 billion in 2007 upgrading wired and wireless network infrastructure. By comparison, all the dotcoms (including Google, Ebay, Yahoo, and many others that signed a letter demanding Net Neutrality regulation in 2009) that are supposedly more important to the Internet economy than the ISPs spent a mere $9.2 billion in 2007 combined! That’s less than the $17.5 billion spent in 2007 by just Verizon building out network infrastructure.
Furthermore, it is a fallacy to suggest that premium services would harm content and application providers because premium peering services are actually cheaper than traditional means. Banning these premium services actually harms the small to medium size dotcoms by preventing them from purchasing cheaper and faster bandwidth. We must not preclude innovative new business models on the Internet and we must preserve the open and competitive bandwidth market because we need the additional business activity and revenue streams that fuels the Internet’s growth.
The myth that dotcoms drive the Internet economy
Free Press is essentially arguing that it is more important to promote dotcom growth and than ISP growth by means of new regulation. But if we look at the numbers, the most dominant dotcom Google needed $29.20 of revenue (2009 data) for every dollar invested on capex. But the three largest U.S. based ISPs AT&T, Verizon, and Comcast on average only needed $6.92 of revenue for every dollar they invested in capex which means these ISPs invest 4.2 times more money for every dollar of revenue than Google.
These capex investments indirectly translate to many more American jobs which means the ISPs are a much larger economic force than the dotcoms. If we look at direct jobs employed by Google versus the three larges ISPs, ISPs employ about 2.9 times more employees for every dollar of revenue than Google. The purpose of pointing out this large disparity is not that we need government policies to promote ISP interests over dotcom interests, but that we need industry neutral policies and avoid promoting dotcom interests over the ISPs.
Effectively killing existing managed services
If eliminating future revenue and investment opportunities weren’t bad enough, Free Press wants to neuter ISP’s existing managed networks and services by disallowing “vertical prioritization” and dedicated bandwidth allocation. Free Press argues:
“Vertical Prioritization” — This model is one where an ISP simply prioritizes its own vertical content and services over all other content. This prioritization can be achieved either by flagging their traffic for priority, or by more subtle ways, such as de-prioritizing applications that are used to deliver classes of content that compete with the ISPs vertical content; or by the outright blocking of an IP application that competes with the ISPs own adjacent market services. Unlike the pay-for-play or pay-for-priority models, this business model involves no new income streams, only the insulation of old streams from network-facilitated competition. But allowing ISPs to insulate their legacy vertical voice and video industries from the natural forces of competition is no recipe for investment — with reduced competition comes reduced investment incentives.
These proposed rules would effectively mean the end of managed services. Even the proposed draft of the FCC NPRM rules preserve vertical prioritization, and rightly so because these managed services such as television and phone service are a significant revenue source for ISPs that provides the critical funding for all the new communication infrastructure. In fact if we simply trimmed the revenues of the three major U.S. ISPs revenue by 25% by severely curtailing their existing video and phone models, and we went by existing ratios of revenue versus investment and jobs, the three major ISPs would have to trim capex investment by $9.7 billion and eliminate 156 thousand jobs. That’s equivalent to the total capex investment of all the dotcoms combined and it is more than the 108 thousand jobs provided by all the dotcoms combined!
Now we would recover some of the investment and jobs if the revenue is transferred to Google television (what YouTube will become) and Google phone (Google Voice) services, but it is an established fact that Google is more “efficient” when it comes to minimizing capex investment and employees while maximizing profit and stock valuations. In fact they’re so efficient that if we project from their current capex spending and employment habits, the transferred revenue would result in Google investing an additional $2.3 billion and employing an additional 37 thousand employees. While that sounds great out of context, it wouldn’t even begin to recover the net economic damage done to the ISP industry. In fact, it would mean a net loss of 119 thousand direct jobs and untold indirect job loss from the reduced capex investments. At a time that unemployment is at an all time high and investment is already weak, the very last thing we need is the kind of Net Neutrality regulations that Free Press is demanding.