One of the pressing questions of the Internet age is whether broadband Internet access is a service so essential that it must be made available to everyone. Should there be truly “universal service” in broadband? At what price? There have been many programs and laws passed on this, but in the last year or so in the US, it’s an issue that’s increasingly come up at the state level. New York has led in this, but other states are also exploring the idea. Light Reading cites a study by the PUC in California’s debate on affordable broadband, and it seems to suggest that affordable broadband wouldn’t have a significant impact on provider revenues. Is that true, does it matter? A lot of that depends on public policy, public attitude, but we can look at the fundamentals to at least get a glimpse of the issues we face.
Broadband policy, like most of public policy, is complicated by the general truth that getting products and services costs money, and that money is easier for some to part with than for others. For decades, it’s been the dominant global policy to establish some form of “universal service”, wherein those who can’t afford the market price for communications are able to obtain it for something they can afford. In recent years in the US, the implementation of this policy has started to shift from imposing a charge on service to create a subsidy pool that then makes up the difference in price, to requiring that ISPs offer broadband packages below market rate to qualified customers. Needless to say, this move has generated pushback from operators who already, in some global markets, are asking for broad subsidies, even for market-priced services.
Those who (like me) are historically minded will realize that what’s happening here is a collision between the established concepts of public utilities and of public-stock corporations. Telecom used to be either a regulated monopoly (a public utility) or even a government depart (postal, telegraph, and telephone or PTT). Today, it’s almost always more like any other stock company, and in fact utilities in general are drifting away from the monopoly model. The question is whether broadband Internet access should be made available at an affordable price, below market price, and if so whether the mechanism to lower prices should be based on subsidies paid to the operators, by operators accepting a lower profit on some customers, or by some combination of the two.
What California is debating is whether operators should offer low-income customers Internet access below the current $30 per month rate. If we look at the report data, it’s focused not on unserved customers, but at low-income customers (those at or below 200% of the federally set poverty line) who are already getting broadband, either at the $30 per month 100/20 Mbps Affordable Connectivity Program rate or by paying market rates for a faster connection. Those groups represent 500,000 and 850,000 customers, respectively. Since the report indicates there are over 5.8 million low-income households, meaning that almost four and a half million households are unserved.
What the report concludes is that the reduction in revenue that would result from mandating a $15 per month service is likely less than one percent, and would be only 2% if all currently served eligible households, including those now electing market-rate plans, were to shift to the new $15 plan. The report goes on to say that many households say that any price at all would deter them from using broadband Internet, and proposes a phased introduction of subsidies from 2026 through 2034, that would eventually address all 5.8 million eligible households. The cost of these subsidies would depend on the mandated operator price of the low-income service tier. The report goes on to say that the benefits to the state of having a broader broadband participation rate could justify these subsidies.
Obviously, taxpayers and voters at any level set “public policy” and I’m not going to comment on what the right policy is, or even if there is one. The technology issue is, IMHO, to frame services, infrastructure, and businesses so that the promised benefits are achieved to the greatest extent, while mitigating any negative impacts as much as possible. Obviously, you can’t mandate profitable operation, and so operators must be able to earn enough to invest, or we’d face a necessary return to a public utility model and perhaps even to a future where broadband services would have to be made a part of the government, as they were in some countries in the past (remember the PTTs?).
To me, this starts with the recognition that universal broadband will require very efficient access infrastructure. Access represents a third of the capex and 42% of opex, according to what I get from operators. To deploy specialized infrastructure to serve low-income households makes it more costly, so as much common infrastructure as possible is essential. This is particularly true if it’s a goal to set relatively high levels of performance on even basic service plans; the 100/20 Mbps service level, for example, is hard to achieve over random telephone twisted-pair.
Both CATV infrastructure and FTTx can provide a framework for delivery of a fairly wide range of services. For example, you can hang an FWA node off a fiber that could also support FTTH, and since many low-income households are concentrated in urban areas where 5G cellular broadband would be efficient, it would be smart to encourage this sort of infrastructure versus trying to make-do with in-ground copper loops.
Opex here may be a larger issue. Operators tell me that, on the average, low-income households are more likely to require support. This is particularly true for senior citizen households, since in general seniors are less tech-savvy. Might it be prudent, as part of an affordable broadband plan, to mandate an AI support chatbot, or at least offer financial/tax encouragement to the use of one? Might a state want to set standards on support chatbots, or even provide a baseline solution that operators would supplement with their own service-specific data?
A similar level of planning is essential in ensuring that the value to the state/government entity created by universal broadband is maximized by planning services delivered over broadband to citizens. Things like broadband telemedicine, for example, are useless if they’re not offered, and if public health programs don’t cover them. All the services likely to be beneficial are also services involving a lot of personal data, and so a mechanism to secure them reliably is critical.
There’s an unfortunate tendency to try to offset real costs with hypothetical benefits. It’s just as dangerous to offset real costs with benefits that you don’t try to maximize, or even assure. I personally think that there really is a benefit to universal broadband, but I don’t think it’s going to fall into our laps. Governments, operators, and enterprises are all subject to economic reality in the end. Dealing with that at a technical level really isn’t difficult, but it is an explicit step that has to be worked for.
