Tag Archives: uc berkeley

Money talks or AT&T broadband walks, CPUC study shows

by Steve Blum • , , , ,

Haas att broadband study

How much money you and your neighbors make determines whether or not you have access to modern broadband service and infrastructure. The network practices study released on Monday by the California Public Utilities Commission cites conclusive evidence of aggressive redlining by AT&T. It is a major – and actionable – report that makes the case against the two companies, but its conclusions come as no surprise.

A study done in 2017 by U.C. Berkeley’s Haas Institute for a Fair and Inclusive Society found that…

The median household income of California communities with access to AT&T’s fiber-to-the-home (FTTH) network is $94,208. This exceeds by $32,297 the $61,911 median household income for all California households in the AT&T wireline footprint.

On the other hand, the median household income of homes served only by AT&T’s 1990s legacy DSL technology is $53,186, according to the Haas Institute. The CPUC study found the same divide between haves and have nots…

Whether deliberate or not, AT&T’s investment policies have tended to favor higher-income communities, and have thus had a disproportionate impact upon the state’s lowest income areas. For example, the weighted average 2010 median annual household income for wire center serving areas that had been upgraded with fiber optic feeder facilities to support broadband services was $72,024, vs. only $60,795 for wire centers without such upgrades Using 2010 US Census data, we find a clear inverse relationship between household income and all of the principal service quality metrics.

The report leaves it up to the reader to decide if AT&T’s income-based redlining is deliberate, but makes it clear that AT&T’s financial strategy is aimed at extracting the maximum dollars possible from communities trapped in legacy monopoly systems…

Persistent disinvestment, extensive affiliate transactions at self-serving transfer prices, extraordinarily large rate increases, and deteriorating service quality all point to “harvesting” as AT&T California’s overarching strategy for its legacy services and customers…The potential gains that AT&T California can realize by raising prices and curtailing investment and maintenance expenditures far exceed any financial penalties it might suffer from persistently poor service quality.

AT&T is not alone. I found a similar pattern in Charter Communications’ investment choices. Until the CPUC forced it to do otherwise, it held low income communities captive in analog-only systems that offered limited television service at exorbitant prices.

California legislature to grant redlining absolution to mobile industry

by Steve Blum • , , , ,

Mobile carriers don’t redline neighborhoods or communities on the basis of income levels. That declaration is the latest present to go under the Senate Bill 649 christmas tree as it nears a final decision in the state legislature. The primary aim of the bill is to give wireless companies open access to street light poles and other “vertical infrastructure” owned by cities and counties in California, at below market rates.

New language tightening up definitions was added to SB 649 in preparation for a floor vote by the California assembly. Some of the bill’s benefits will only be available to licensed mobile carriers, rather than any wireless Internet service provider that comes along, and it further narrows the ability of local governments to restrict cell site construction on aesthetic grounds.

To justify these gifts to the mobile industry, lawmakers included language praising mobile carriers…

The Legislature further finds and declares that wireless service providers deploy small cells to areas based on demand for services regardless of the income characteristics of the areas, that this act will complement efforts to close the digital divide by creating a framework that will incentivize private industry to invest or accelerate investment in the deployment of small cells, and that this act will complement current state and federal government efforts to subsidize the deployment of broadband.

There are a couple of problems with this statement. It’s microeconomic nonsense to say that demand isn’t connected to income levels. So if lower income levels lead to lower demand in a given area, then mobile carriers won’t improve service: less income = less demand = less infrastructure investment.

It also ignores the fact that, in addition to being a mobile carrier, AT&T is also by far the largest traditional telco in California and, according to a U.C. Berkeley study, it does redline wireline customers in poorer communities. AT&T’s solution is to replace ageing wireline networks with low capacity wireless systems, while upgrading to fiber in high potential neighborhoods. That will be a digital divide that the California legislature will never be able to close, even if it wanted to try.

Consider who pays for broadband studies, but don’t stop there

by Steve Blum • , , , ,

Gamblers or exiled royalty?

I’ve commented on a couple of university studies recently, one critical of municipal broadband’s business model and the other ripping AT&T’s infrastructure upgrade redlining in California. In neither case did I write about who picked up the tab for the work. That’s because I thought that both analyses stood on their own. But it’s a fair question to ask and, for the muni broadband study at least, it’s a significant one because the source of the money was the primary basis for challenging the work.

The AT&T study was done by U.C. Berkeley’s Haas Institute for a Fair and Inclusive Society, but was largely paid for by the Communications Workers of America, which is the primary union representing the company’s employees. And which was in the middle of a major contract dispute at the time.

The University of Pennsylvania’s report was published by its Center for Innovation, Technology and Competition, which gets its funding from several self-interested contributors, such as AT&T, Comcast, Verizon and lobbying fronts for the cable and mobile industries. But there’s also representation on the list from players who often end up on the other side of the table, including Facebook, Microsoft and the Internet Society.

But that’s the way the system works in academia, as The Economist notes

Derek Bok, a former president of Harvard, once observed that “universities share one characteristic with compulsive gamblers and exiled royalty: there is never enough money to satisfy their desires.” This is a bit hard on compulsive gamblers and exiled royals.

The way to challenge an analysis done by a reputable institution – and I’m not generally including Washington think tanks in that category – is on the basis of the methodology and data used. In both studies, the method was solid. I can’t fault the Haas Institute’s work – they ran the same kind of analysis on CPUC and FCC data that I regularly do, and came to very similar conclusions. Likewise, there was no reason to fault the methodology in the Pennsylvania report. It was pretty basic comparative analysis of financial results.

There were problems with the data in both studies, but that’s not the fault of the authors. Muni fiber to the home results are not published with the same rigor as those from publicly traded companies, if they’re published at all. The CPUC data that the Haas study looked at is only so granular – it gives a pixelated view of broadband availability at the census block level, but no better.

AT&T fiber redlines low income communities, U.C. Berkeley study finds

by Steve Blum • , , ,

Where high income households are thick on the ground, AT&T builds out fiber to the home systems, but does minimal upgrades for middle income areas and leaves low income communities with 1990s-style legacy DSL or nothing at all. That’s the top line conclusion from a study done by U.C. Berkeley’s Haas Institute for a Fair and Inclusive Society

  • The median household income of California communities with access to AT&T’s fiber-to-the-home (FTTH) network is $94,208. This exceeds by $32,297 the $61,911 median household income for all California households in the AT&T wireline footprint.
  • In contrast, the median household income of California communities for whom the most advanced broadband technology available from AT&T is its slower U-verse fiber-to-the-neighborhood (FTTN) network is $67,021, which is $27,187 (28.9 percent) lower than the median household income of fiber-to-the-home households.
  • Approximately one-quarter (27.6 percent) of households — about 2.7 million households —in AT&T’s California footprint are stuck with slow DSL. The median household income for California households for whom DSL is the most advanced broadband technology available from AT&T is $53,186, which is $41,022 (43.5 percent) lower than the median household income of fiber-to-the-home households.

There’s also a distinct urban/rural divide in AT&T’s broadband infrastructure deployment strategy. While metropolitan areas get fiber and VDSL upgrades, rural areas are ignored. According to the study, almost no homes in 14 rural counties have access to AT&T broadband at the FCC’s minimum standards of 25 Mbps download/3 Mbps upload speeds and one-third lack access at the CPUC’s minimum of 6 Mbps down/1.5 Mbps up. In its overall service territory in California, 252,000 homes do not have access to AT&T broadband service at all.

In many respects, the report’s findings are no surprise. AT&T has been very clear that fiber infrastructure would only be going into high potential areas and that it plans to rip out copper networks in rural California and replace them with wireless service.

The study recommends that policymakers, and the California legislature in particular, should demand greater accountability from AT&T and promote more equitable high speed broadband deployment. Unfortunately, the California assembly has not taken the study’s findings or recommendations to heart. It just voted to lower California’s minimum broadband speeds, specifically to accommodate the substandard technology that AT&T maintains in rural and lower income communities.