Editor’s note: This commentary is by Jock Gill of Peacham, an internet communications consultant who served in President Bill Clinton’s Office of Media Affairs. He is town energy coordinator in Peacham.
Could Vermont solve its known broadband inequality issues in a timely and cost-effective manner by leveraging SpaceX’s Starlink system? I note that North Carolina is already exploring this option. See: https://www.tesmanian.com/blogs/tesmanian-blog/nc-starlink
As we think about broadband solutions, we need to ask 1] what is the cost per connection? 2] When could all connections be made? 3] What is the cost per month to the end user? 4] Should broadband be infrastructure or a billable service?
I am suggesting that we look into making it very attractive to Starlink to turn on its service in Vermont in general, and the Northeast Kingdom in particular, sooner. Could we achieve this goal if we guarantee a reasonable number of takers? The state published estimates in May that about 80,000 locations have inadequate or no access to broadband. How many locations would we have to offer to be attractive to Starlink as a reasonable test?
The NEK has around 20,000 locations that need a good enough broadband solution. A connection to Starlink may be generally available to the residents in the 30 towns that are members of NEK Broadband for about $400 per location. There is simply no way that fiber, perhaps the goal for an ideal solution, could reach all NEK locations for $400 each. Much less in less than 12 to 18 months. Perhaps even sooner, if Starlink is willing to cooperate. NEK Broadband would be a logical organization to manage a relationship with Starlink for the benefit of its members.
I am, for the record, not talking about investing money in Starlink, a privately held company. I am simply suggesting that Starlink may well be a good enough as well as an affordable and timely solution. We must not allow the perfect to be the enemy of the good enough. Not if we want an affordable solution in the nearish term.
The simple question is this: if the state reached out to Starlink and offered to install X receivers in the NEK, what would Starlink say about price and availability?
Let’s do the math. It is thought that the Starlink satellite solution will cost Starlink $7.2 billion to build out. It has further been suggested that the fiber solution that some prefer will be one-twentieth of that cost. Thus the fiber solution is about $360 million. Hopefully, this would cover the entire state. Assume 80,000 locations need help; then this is a cost to Vermont of roughly $4,500 per location.
On the other hand, a Starlink connection would be about $400 per location, one-tenth of the fiber solution to the problem. This is my point. Why not leverage Starlink’s investments to our advantage?
Further, we must consider the per-month costs to the end user for the ideal fiber solution. If we look at ECFiber’s pricing, it is not very affordable — except possibly at the slowest speeds. There is simply no point in providing a Cadillac fiber product to users who are not Cadillac customers. As an aside, there is no need for tiered pricing, as bits are infinite, free and have zero cost to anyone.
Lastly, we have the question of whether broadband should be considered infrastructure or a billable service. Take the case of people living in a trailer park. They are not charged per mile they drive on a road, nor are they charged for the speed at which they drive. Why should driving on the internet, just as vital today as road infrastructure, be any different? This argues for broadband to be treated as infrastructure, owned locally — just as many of our roads are town roads.
As Christine Hallquist says: “Starlink is a fine interim solution.” We should look into it. It would be financially irresponsible not to.
In the end, we have to take a “show me, don’t tell me” approach to all suggested broadband solutions.
