The California Public Utilities Commission, on the other hand, decided to go in the opposition direction and unanimously endorsed the higher standard of 25 Mbps down/3 Mbps up yesterday. That’ll have no effect on AT&T’s fixed wireless roll out though, whenever that actually happens.
There seems to be a different between making it official and making it real. Using the link in the press announcement, I checked a dozen locations in California where AT&T has claimed federal Connect America Fund subsidies – where its wireless local loop service is targeted – and all came back with “AT&T fixed wireless Internet isn’t in your area yet”. I got the same response when I entered the zip codes for a couple of the Texan counties that AT&T specifically called out as ready for fixed wireless service in a separate press release.
California was on a list of nine new states, bring the total where AT&T claims to offer its 10 Mbps down/1 Mbps up wireless substitute service to 18 states.
According to Ars Technica, the base rate for the service is $70 a month, or $60 a month with a contract. AT&T isn’t disclosing the monthly rate on its website, but it does helpfully point out that the base service only includes 160 gigabytes a month. Anything over that costs $10 per 50 GB, up to $200, for a total max charge of $270 a month.
Pricing and service levels are reasonably clear though, at least as clear as such things ever are. If you want the nominal 10 Mbps download/1 Mbps upload speeds that AT&T is required to provide where it’s getting federal Connect America Fund (CAF) money, the rack rate is $99 to have it installed and then $70 a month for the first 160 GB of data. Every 50 GB (or fraction thereof) of data over that cap will cost an extra $10, up to a a maximum overage charge of $200.
It’s costly bandwidth, compared to the standalone ADSL2-based wireline service I get from AT&T. I pay $57 a month for “up to” 12 Mbps download speed, with a 1,024 GB cap and no currently defined upload speed. Do the math: customers getting the federally subsidised wireless service would have to pay $250 per month for the same amount of data I get, at a lower advertised speed.
Signing up for bundled DirecTv service or for a term contract can bring the monthly price down, but that’s true for wireline subscribers too.
There’s also reason to question whether AT&T will actually deliver on the required 10 Mbps down/1 Mbps up service level. As is typical, the AT&T web page touting the wireless service qualifies its promise of “at least” 10/1 by adding “data speeds can vary depending upon various factors”. Right. Like trees, weather, and your neighbors streaming House of Cards on Netflix.
AT&T and Frontier Communications have scooped up the CAF money on offer in California. Frontier hasn’t provided any information yet on how or if its federally subsidised service will differ from the norm, but it’s clear that AT&T is using the money to lock rural Californians into service – for decades to come – that’s slower and more expensive than what their urban and suburban cousins enjoy.
In 2016 (which is the benchmark year for 2017 rates), urban customers subscribing to packages with download speeds of 10 Mbps, upload speeds of 1 Mbps per second and a data cap of 100 gigabytes per month – in other words, the slowest and lowest service – paid $76.49 per month. That’s $7.33 more than a year before, an 11% increase. Customers with the highest end service in the survey – 25 Mbps down, 5 Mbps up and no data cap – saw their bills go up only $1.52, increasing 2% to $90.76.
As the table above shows, the lower the level of service you buy, the greater the price increase you have to bear, both on an absolute and percentage basis.
One caveat: the benchmarks are based on the prices and terms that are offered by Internet service providers, and not on the average price that consumers actually pay. In other words, customers with low end packages might be – probably are – paying less on average than the benchmark price because when presented with a choice of comparable packages, the microeconomic assumption is that they’ll opt for the cheaper one.
Urban and suburban residents in California can typically – but not always – choose between service from a cable and a telephone company, for example, so they can make that choice. On the other hand, the benchmark rate, which also factors in expensive fixed wireless prices, would be the best that many rural residents might be able to get from the single, federally subsidised provider that serves their area.
Who would have thought that the Federal Communications Commission’s first significant decision of the Trump era would be to take money originally designated for its no-incumbent-left-behind broadband subsidy program – Connect America Fund 2 (CAF-2) – and use it to top up reasonably competitive state grants, with the state calling the shots?
The $170 million that would have gone to Verizon will go to winning applicants for New York’s broadband infrastructure subsidy program. The federal money will be given as a one to one match for state subsidies awarded in CAF-2 eligible census blocks, but only up to the maximum amount that the FCC originally designated for those blocks.
We find that New York is uniquely situated to quickly and efficiently further our goal of broadband deployment. New York has committed a significant amount of its own support—at least $200 million—to…its broadband program that is designed to be compatible with and achieve the goals of Connect America Phase II. Moreover, New York is poised to quickly implement the next phase of its program in a matter of months so that deployment of broadband of speeds that meet or exceed the Commission’s baseline requirements for Connect America can be achieved while the Commission is in the process of finalizing and implementing the Connect America Phase II auction.
Virtually all the CAF-2 money that the FCC offered in California was accepted (CenturyLink turned down $300,000 to serve 45 homes in Modoc County), so the New York decision can’t be cloned here. But the FCC has shown a new willingness to work with effective state broadband programs, rather than at cross purposes. These are interesting times.
There are solid engineering arguments why 700 MHz is more valuable than 2,300 MHz for delivering wireless broadband to homes. I’ll address only the line of sight argument. The larger-wavelength 700 MHz frequency penetrates buildings and foliage better than the shorter-wavelength 2,300 MHz frequency. For 2,300 MHz service to reliably deliver broadband, there needs to be line of sight, meaning you can physically see, without obstruction, between cell tower antenna and customer antenna. Without line of sight, service will be less reliable than with 700 MHz. Many parts of rural California are mountainous and have trees, which makes line of sight broadband difficult and expensive to deploy.
In terms of performance, 2.3 GHz is pretty much the same as the 2.4 GHz band used for WiFi. AT&T plans to install outdoor antennas and receivers at customers’ homes, which are absolutely necessary in order to get through walls that block those frequencies. But foliage, particularly when it’s as dense as it is in the mountain areas of California, can stop it just as cold.
In theory, to qualify for the federal money, AT&T has to reliably serve virtually all the homes in the subsidised areas with 10 Mbps download and 1 Mbps upload speeds. One thing to watch for: the time of year AT&T does its testing. What’s fine at 2.3 GHz in the dead, grey winter won’t necessarily work when summer comes and everything turns to green.
AT&T’s rural California road show is continuing, as the company pitches county supervisors on the wonders of wireless service and the need for speedy approval of towers and other infrastructure, without making it clear that the plan is to use it to replace copper wire networks.
AT&T is now looking to bring high-speed internet service to underserved areas in Humboldt County, which will require the company to construct new cell towers. [AT&T staff lobbyist Marc] Blakeman said they are looking for landlords, including the county, that are willing to allow them to build new towers near Fieldbrook, Blue Lake, Jacoby Creek, Kneeland, Bridgeville, Redcrest and Miranda.
As they are using federal funds for this project, AT&T must meet certain deadlines such as improving service to 40 percent of underserved areas by the end of next year.
What the story didn’t pick up on is the fact the federal money referred to – the latest round of the Federal Communications Commission’s Connect America Fund program – is intended to boost residential, and not mobile, service. AT&T claims to be doing that too, but glosses over the fact that these so-called residential upgrades will come via its still mysterious wireless local loop (WLL) technology, which will be grafted onto mobile infrastructure.
AT&T claims that WLL will meet the federal broadband service benchmark of 10 Mbps download and 1 Mbps upload speeds, and is capable of even faster throughput. What’s not known is how much capacity a given WLL access point will have – in other words, how many homes will actually be able to dependably get those speeds.
Naturally, AT&T isn’t mentioning the serious weaknesses of its plan to replace copper service with WLL and, unfortunately, few people at the local level are questioning the happy happy, joy joy presentations offered by its lobbyists.
The video was streamed live, and I’ve posted a recording to YouTube. It begins a short time into the presentation, because the county’s feed began while it was in progress. Not much is missing – a few seconds maybe, but not more than a minute or so.
In the first rollout presentation that I’ve watched of what AT&T has called wireless local loop technology, Perez said that it’ll support 10 Mbps download and 1 Mbps upload speeds, which meets the federal CAF standard, but fails to meet California’s minimum standard of 1.5 Mbps on the upload side. And it falls far short of the federal advanced services minimum of 25 Mbps download and 3 Mbps upload speeds.
She also alluded to replacing wireline telephone service with mobile phone technology – “VoLTE-based telephone service”, as one of her slides put it.
Consistent with AT&T’s past descriptions of WLL service, Perez positioned it as a bolt-on to existing cell sites, which is presumably where the voice over LTE based phone service would come from. In response to a question from supervisor Shiva Frentzen, Perez claimed that “this is not fiber funding”. That’s not true: CAF subsidies can be spent on any technology capable of meeting the 10 down/1 up standard. If AT&T isn’t spending it on fiber, it’s by choice.
Actually, even in rural California, I expect some of the money will go toward fiber back haul connections to cell towers, particularly the new ones that Perez said would be necessary. Her pitch ended with a plea for rapid approval of permits for the work, which met with a generally sympathetic response from supervisors and staff.
Nationwide, the Federal Communications Commission has a total of 1.5 million homes and businesses on its preliminary eligibility list. About two-thirds of the Californian locations are in what are called extremely high cost areas, which means that the estimated per location subsidy necessary to convince a telco to build out broadband infrastructure is more than $1,200, according to the FCC’s funding model.
The upgrade cost as calculated by that model, less the $300 per location that the FCC thinks telcos should be spending regardless, will be the starting price for any particular census block. Bidders will make offers, presumably at progressively lower levels, until the FCC has decided that it’s getting the biggest bang for its billion or so bucks. How that decision will be reached is still to be determined.
If you add up the starting prices for all 1.5 million locations, the total comes to $5.8 billion. That’s more than four time the available cash, so there’s likely to be a lot of homes and businesses that’ll be left out. Unless bidders are feeling particularly generous, but I wouldn’t bet on that.
The total starting price for the locations in California is $434 million, which is about a thousand bucks per location more than the national average. Even if the FCC factors speed levels and cost differentials into its decision, in addition to the raw number of homes and businesses, Californian ISPs will have to shave more off their bids than those in most other states.
In comments it filed regarding the FCC’s proposed bidding rules for the next round of CAF-2 subsidies, Verizon wants extra credit given if it makes minimum service – 10 Mbps down/1 Mbps up – bids for unserved areas it turned down last year, but doesn’t want competing bids to be given greater weight if higher, even gigabit, speeds are offered, because, well, who needs all that bandwidth…
For every location in the gigabit tier that is awarded support because of a large weight, several other eligible locations would be left without any broadband service whatsoever. These customers would have been well served by services offering the “baseline” speeds and capabilities, which are sufficiently robust to support most consumers’ online activities.
What’s doubly disingenuous about Verizon’s comments is that it’s trying to game the system so that it can cherrypick the rural communities in its service territory that it deems sufficiently lucrative to upgrade. In the first round of CAF-2 subsidies last year, Verizon could have had those subsidies handed to it on a platter, if it agreed to upgrade all eligible locations in a given state.
Verizon said no, preferring instead to come back later and bid on the particular areas it wants to serve. As it’s entitled to do. But the FCC shouldn’t let Verizon use its existing monopoly advantage to all but automatically claim cherrypicked communities and lock them into the lowest tier of service, while ignoring the rest.