Can broadband networks be made both populist and profitable? What constraints on technology and capability exist in meeting that goal? Obviously this is one of those things that require a balance between the characteristics of a service and the cost, that balance being impacted by the distribution of the service users and thus the efficiency of infrastructure required.
Broadly speaking, the ability of a market geography to provide a return on telecommunications service is a function of two things, which I’ve modeled for decades and call “demand density” and “access efficiency”. Demand density is the gross domestic product (GDP) of a geography, per unit area. Access efficiency is a measure of the distance from service users to a point of pubic right of way, normalized to the state of last-mile technology effectiveness.
My work on this has, for convenience and based that as a US resident I have better access to US data, digested all of this into a single number which is normalized to the US value (the US has a “demand density” of 1.00). Broadband pays back better than the US where demand density is higher (most of Europe, Japan, Singapore), and is more challenging where it’s lower (Australia, Canada, Africa, Central and South America). Generally, demand density is high in first-world economies with high GDPs and a high level of urbanization.
In almost all countries, there are variations in demand density through the country, based on population density. It’s possible to derive regional values where there is regional economic data (like GDP) available, but to get the information more locally is difficult because the relationship between service opportunity and some measurable economic factor is more difficult to obtain. It also more difficult where a country has a highly urbanized population but the population is concentrated in a limited geography, which means the opportunities in non-urban locations are highly specialized. A single community in a ten-thousand-square-mile area can still present a good opportunity if the population is high/well-off, but distributing the population equally over that area would surely make the ROI on service deployment very low.
Obviously, universal broadband has to deal with this overall situation, so how do we relate that goal to the concepts of demand density and access efficiency? It’s somewhat complex, somewhat sketchy, but illuminating.
Demand density, as my single normalized metric, can be low for three reasons. First, the economic power of the area can be small. Second, the geography can be large. Third, the area may not be readily amenable to deployment. All of these cause a high-level problem, but each can require a different solution.
Low demand density due entirely to economic limitations is a subsidy problem. The presumption we can make is that infrastructure is “normally” efficient but that normal infrastructure efficiency and service cost don’t align with willingness/capability to pay. The decision to address this is pure public policy and politics.
Low demand density due entirely to a highly dispersed population is at one level also a public policy question, but one most can answer. If you choose to live off the grid, then you either have to bear the cost of being connected yourself, or accept not being effectively connected. A billionaire who decides to build on a mountain top can’t run to taxpayers to pay for fiber to run up to his home. Low demand density due to access efficiency issues alone is similar. An island community may well have to expect to pay as a community to link to the online world.
The challenge, then lies in the fuzzy area remaining, and in the fact that most “universal broadband” strategies really are only subsidy strategies. A subsidy approach to anything other than our first economic reason for low demand density is going to promote abuse in one area, and be insufficient if low economic power combines with a service geography that can’t be efficient. Put a lower-economic quintile person on the mountain top. A subsidy won’t encourage the local fiber access provider to trench up to them. Thus, you need a universal broadband policy that makes some policy and technology decisions, either directly or in the form of explicit targets based on what population you’re trying to support.
The urban poor are reasonable subsidy targets. The rural, even deep-rural, poor are less so, because a subsidy makes a marginal ROI viable, but doesn’t address the problem when the cost of deployment is outlandish relative to the empowered population. What’s the answer? RF in the form of FWA and satellite broadband.
My research shows that 85% of the US population lives within the economically viable footprint of current access technology, which means that it could be reasonable to offer subsidies to operators to cover any under-served population in these areas. The remainder live largely outside that area, and thus could likely require something non-traditional in access technology. For the “edge” of the 85% area, FWA could be a solution, and my estimate is that 8% of the remaining 15% of the population could be addressed that way. For the remainder, only satellite broadband is likely to be adequate.
For the 15%, access efficiency details really matter. Generally, access efficiency is highest where there are well-used public roads whose rights of way can be leveraged for cabling, towers, etc. These also tend to concentrate households and businesses, which creates better local demand density. Where access efficiency is low, satellite is likely the only option.
Part of the problem is that there are two constituencies pushing subsidization. One is the vendor community, who often push blunt subsidy programs like BEAD (Broadband Equity Access and Deployment) that have an arguable focus on fiber broadband, and the other a public policy advocate community that tend to push programs that target populations without any technology focus at all. As the referenced article shows, BEAD is being criticized for fiber focus, and in fact there is advocacy for the very FWA and satellite focus I’ve talked about here. The question is whether any form of subsidy can really indirectly target and promote a technology by targeting the buyer of the service alone. I don’t think so.
There are ways to promote “universal broadband technology”. Spectrum auctions could offer refunds for meeting FWA goals, and the same for satellite providers. You could offer rights of way, launch benefits, and so forth, too. At the minimum, you’d need to redefine “voice” and “text” to a higher level, so that satellite or traditional networks could serve both, and so that connections could be preserved across the two network options for roaming users. Today’s practice of either requiring a different mobile device, or supporting emergency texts as an option via what’s essentially a separate device feature, isn’t optimum.
Per-user subsidies are a public policy decision that’s as complex as any sort of subsidy decision, but subsidies that require or imply a service modification are way more complex than political processes seem capable of dealing with. We have that here, and until we either figure out a way to make political/technical processes work in both dimensions, or admit to failure and let market conditions take their natural course, the universal broadband initiatives we undertake are almost surely doomed.
