The Computer & Communication Industry Association (CCIA) released a paper recently warning about the impact of imposing Universal Service fees on what it characterizes as cloud services. CCIA is an association that lobbies on behalf of some of the largest web companies like Amazon, Meta, Google, Apple, Netflix, and Cloudflare.
The author quantifies the impacts of imposing a USF fee on cloud service providers and uses an example of a 5% USF fee on cloud services. Some of the cited impacts include:
- The author of the report says that the cloud service industry is extremely price-sensitive and that there would be a loss of business if a cloud service provider passes USF fees to customers. The report says that a 1% increase in the price of cloud services would result in a drop in business between 0.5% and 0.6%.
- A 5% fee passed on to customers would decrease national GDP between $59 billion and $148 billion annually. This means impacts on State GDPs, with California having a downside as large as $25 billion, Texas an impact up to $17 billion, and South Carolina an impact of as much as $2.2 billion. To put that last number into perspective, the total annual State budget for South Carolina in 2024 was only $40.2 billion.
- The industry invests around $54 billion annually on infrastructure, and the report suggests that this would cause the industry to cut back on new investments by $7.6 billion.
- A 5% fee would increase overall national inflation by up to 0.13%.
It’s somewhat refreshing to see that the lobbyists in other industries are as willing to greatly exaggerate claims about the impact of regulatory and legal changes as the telecom industry. Let’s put the claims to a reasonableness test.
The total Universal Service Fund was $8.5 billion in 2024. USF is funded today entirely by a fee levied on telecommunications services. There are no proposals that I’ve seen that would ask cloud companies to fund that entire amount – the current bills being contemplated by the House and Senate would ‘equitably’ allocate the fee between telecom services, broadband services, and ‘edge services’, which is the same as what the paper calls cloud services.
It’s hard to think that the share of USF allocated to cloud companies would be much more than perhaps $3 billion per year – far smaller than the 5% increase included in the CCIA analysis. A $3 billion fee would not create the dire consequences warned by the paper.
Cloud companies are already paying a significant portion of USF fees that are imposed on the fiber circuits sold for Internet backbone. Fiber circuits to data centers are a big piece of the fiber transport network that is part of the telecom base for USF. Reducing the USF fees paid by telecom companies would significantly reduce the fees charged for backbone fiber. This might offset a significant portion of new direct fees on cloud companies.
The interesting thing that nobody knows yet is which companies in the cloud industry would pay any new fees. That’s something that would only be solidified by Congress at the time they adopted the new method of funding the USF. The fees charge to telecom companies are assessed on everyone from the giant telcos down to the smallest telco. The impact of a USF fee on cloud companies will be diluted if the fees are spread across a wider number of companies than just the large members of CCIA.
The bottom line is that the CCIA analysis is downright silly when it threatens that a $3 billion fee on cloud services might wipe out 5% of South Carolina’s annual State budget or might decrease national GDP by at least $59 billion. But you have to give it to lobbyists – the purpose of the analysis was to provide talking points for politicians in DC, and I suspect it provided the talking points CCIA was looking for.
This blog is being published on Friday morning. Later this morning, the Supreme Court should be announcing the results of the case that asks if the FCC has the authority to operate the Universal Service Fund. Expect my reaction to that case on Monday if the Court decides the FCC has the authority – expect something sooner if they don’t.
some factual issues here. USF fees are for end users and are paid based on the price of the service. The bulk of this is residential home users followed by end-use business services. The rate is a whopping 29%
Cloud providers pay an incredibly small percent because there are no USF fees on settlement-free peering, any private peering, transit cost. Amazon pays no USF fees on their transit between their datacenters and no USF in their peering with providers. You might find loose change somewhere that cloud providers are paying USF but in all practical terms they are not.
The cloud hosting market is already showing signs of pricing issues. Seems like every few weeks a very large company announces that they’ve left the cloud. Price pressure here is very high. Any extra fees likely only benefit Amazon, google, and MS who get’s selected for reliability/infrastructure over price. It likely hurts smaller providers a lot. And I’d argue that .5% losses stated are likely yearly losses so a compounding hit.
The idea that adding taxes and fees to things wont have an affect to sales volume is just crazy. Consumers are stretched pretty thin, as is evidenced buy the rapid rise of FWA from cellular vendors who sell a clearly inferior product, but at a price point more palletable. Adding USF fees to cloud companies will increase prices on essentially all services hosted ‘in the cloud’. expect price hikes to prime, netflix, hulu etc to follow this putting pressure on consumers. And hosting price pressure will certainly push more companies back to their own datacenters.
I’m not saying not to do it, I’m arging against the idea that it’s a minor thing. I don’t know that cloud services are better than self hosted services. I run both, there are advantages both ways. For small companies the cloud reduces labor because you likely don’t ‘need’ a full rack of servers and the corrisponding maintenances etc. So this could be a fee that leans heavy on smaller cloud operators and smaller businesses as well.
Could mean the ‘return’ of the server admin though. on-prem servers need on-prem management. And it’d be fair to argue that the cloud services, amazon in particular, and basically running on tax liability loopholes and don’t pay taxes that corrispond to their massive earnings.
I think it’d be way more appropriate to address the taxing of profits, hiding of costs, etc than to try to extend USF.
The CCIA made an assumption that there would be a huge tax, in the order of tens of billions, assessed to cloud companies. The bills in Congress don’t talk about charging cloud providers, but talk about charging ‘edge providers’ to support USF, which means the many large companies that use the Internet. Most of those fees would also be passed on to end users. Fees to Netflix would likely be tacked into subscriptions. Fees to Google might get tacked onto the fees to advertise in the search engine. Amazon has hundreds of different revenue streams and could tack fees on product lines other than AWS.
I’m not necessarily in favor of this, but the fees have to be spread somewhere else if the FCC is going to keep operating the USF. A fee on broadband service in addition to telephone service might be enough. The ISP and edge industries are going to be slugging that battle out, for sure.