Hacker News Reader

Comment by liquidise | Hacker News Reader

This headline is only part of the story, looks like ~10 ISP's were awarded sizable subsidies.

While i am not a fan of subsidies in general, the ISP space is particularly tricky for free-market forces to ensure competition. The capital requirements are overwhelming in many cases for a traditional startups, leaving the only other option as local/municipal votes, which have been successful in some areas.

Long-term, i could see internet access to receive utility status in the states. Until that happens, funding ISP competition, particularly in rural areas feels like a wise move.

View on HN

> The capital requirements are overwhelming in many cases for a traditional startups, leaving the only other option as local/municipal votes, which have been successful in some areas.

I disagree. Especially today, capital is highly abundant and dirt cheap. The markets are starved for yield on any sort of fixed income product, especially when it's secured by tangible assets with predictable cash flows. Every day the market sees new successful covenant-lite bond and loan sales raised with payback periods measured in decades.

The larger issue is that many municipalities explicitly enforce the local cable co's monopoly. This takes the form of exclusive "pole attachment" contracts. The local governments took kickbacks from the first broadband company that came along. For a quick payday upfront, they screwed their constituents out of any sort of competitive marketplace.

> I disagree. Especially today, capital is highly abundant and dirt cheap. The markets are starved for yield on any sort of fixed income product, especially when it's secured by tangible assets with predictable cash flows.

I disagree here. Capital might be highly abundant and cheap in general but not necessarily in the ISP sector for startups. There are enormous costs associated with laying and maintaining fiber lines or building/launching internet satellites. SpaceX has an advantage by being able to launch their own satellites, it might not be profitable if they had to contract out all of their satellite deployments. I don't think investors will be eager to throw lots of capital at an ISP startup, especially after Google Fiber.

The "last-mile" problem has been around for a long time and doesn't seem to be going away, although wireless solutions (including Starlink) is one way to bypass that problem.

My recent employer was one of the ISP providing rural telecom service. These subsidies accounted for a hefty percentage of our income and the company was still perpetually on the edge of failure. Most of our competitors were in similar situations, loaded down with breathtakingly large amounts of debt and expensive to maintain decaying infrastructure, much of it purchased from the Baby Bells when they got out of the market. I can't say that the company was particularly well run but am certain that without the government subsidies, the business would collapse and large numbers of rural customers would be without telecom services, which extend beyond just internet access.

Not sure what Starlink's existence is going to do to the company. Offering 25 Mbps aDSL with low reliability isn't going to compete very well if a customer can get 100 Mbsp service for a similar price and higher reliability. On the other hand, this might be justification the company can use to stop servicing the most expensive of the rural accounts to service while focusing on small town gigabit service that can compete with a reasonable expense to service.

Isn't Starlink supposed to have the lowest ping of any solution out there (light not travelling in fiber but air/space) and be a breakthrough for financial customers? Bandwidth doesn't really matter _too_ much in those applications does it?

According to Morgan Stanley (not the final arbiter) Starlink is the bulk of SpaceX's commercial value: https://www.cnbc.com/2020/10/22/morgan-stanley-spacex-to-be-...

SpaceX currently has a massive cost advantage in replenishing the satellite fleet which ever advancing equipment.

It already offers the lowest transcontinental latencies which for high frequency trading alone would be a bonanza.

>It already offers the lowest transcontinental latencies which for high frequency trading alone would be a bonanza.

Inter-satellite optical links aren't ready yet (and they're only barely getting to MVP on their initial generation sat shell). When those come online yes, Starlink should have a latency advantage over most any distance sufficiently long enough to make up their fixed RTT mileage cost with their 40% speed advantage vs standard optical fiber [1]. But for the current and near future deployment Starlink satellites act purely as "bent pipes" connecting terminals to a groundstation, both with LOS to the sat.

HFT gets cited a lot but probably one of their biggest bonanzas once intersat links are up will be marine and aviation internet. If you think rural pricing is bad...

----

1: Possibly with a very few rarified exceptions that use terrestrial microwave links or if someone went to the trouble of using photonic bandgap fiber.

> one of their biggest bonanzas once intersat links are up will be marine and aviation internet. If you think rural pricing is bad...

I'm just not convinced about marine - IMO, there just aren't that many customers out there.

Aviation could be interesting, however, post Covid.

One market you didn't mention is military. I can see exactly why the military is cozying up to SpaceX over Starlink. They've had multiple failed attempts[1] at integrated networked warfare initiatives. Starlink seems like a backbone the military could use to implement their vision piecemeal (and with probably a much higher success rate).

___

1. One example: https://en.wikipedia.org/wiki/Future_Combat_Systems

I'd love to get 25Mb DSL but you can't get that in the US where you're lucky to have anything above 4Mb. Not all of us need 100Mb+ internet access.

I live in the midwestern US and I can get 80/20 Mbps VDSL from CenturyLink. Obviously results will vary depending on your exact location, but to say "you can't get [25Mb DSL] in the US" simply isn't true.

The infrastructure those companies have built already exists, and the subsidies are mostly going to pay debt. If the subsidies stopped, the companies would go bankrupt. Their assets would be sold to another company, likely at a price less than the debt. The new company could therefore have a lower debt burden, and wouldn't necessarily need the subsidies to provide the same level of service. The subsidies are stopping the market from efficiently allocating capital.

Sure, but what about the people who will have disrupted or potentially no internet access/support while the market is sorting itself out?

If your plan is to be carried out, then there is ripe opportunity for a politician to come to these areas an say, "vote for me, I will protect your internet access" and then fight for the funding instead.

Treat it like any other utility - it continues to run, just without servicing its debts. Most of the expense is paying for capital. Stop that (because bankruptcy) and the ISP can be profitable, and sold to whoever pays the most.

There is some similar thing that happens with ski resorts. Something like 3 bankruptcies and debt buyouts before they are successful.

But why is it a problem that rural areas don't have telecom services? If they truly valued those services, they could move to where they are available.

These ISP subsidies are a classic example of government waste. ISPs promise, don't deliver, and are awarded even more cash to burn. There's no reason we should support this waste of our tax dollars.

Because the government has decided that the existence and viability of rural communities is a policy goal in and of itself; and that basic telecommunications services are necessary for public safety and good government in such communities.

See also USPS's Universal Service Obligation, which requires it to have a minimum level of service to every single address/user in the US.

This is common in developed countries (e.g. the EU requires all its member states to have universal service for both post and telecommunications). China isn't there yet, but is working towards that goal.

Similarly, most developed countries have some form of universal healthcare.

Nonetheless in America, this is considered an un-American example of socialism run amok. The same pro-individualist criticism applies to rural telecom, postal service, and infrastructure. And it applies even more directly to rural hospitals.

It does seem wise, except I believe that ISPs have already received large subsidies in the past and then never completed their end of the bargain. Without consequence, of course. I hope none of those companies received any more subsidy.

> never completed their end of the bargain. Without consequence, of course

That's exactly why they need more money!! Because, maybe, this time they will?

If not, give them a bit more!!

Government subsidies at their finest! /s

For financial comparison, New Zealand (similar size/population to Oregon) used a lending model to fund a national fibre rollout in a private-public partnership.

“The Government will [initially] invest up to $1.35 billion [NZD] in the UFB network, which it expects to cost a total of $3.5b [NZD]. That includes $929m which it will invest directly in Chorus in interest-free loans and equity. [The government] will not be entitled to any dividends until 2025 and the loans will not be repaid in full until 2036. The difference between what it will pay and what it will get back will be $600m in today's money” - https://archive.vn/20170807221541/http://www.stuff.co.nz/bus...

84% of NZers can access fibre, and uptake is 64% as per: https://www.mbie.govt.nz/assets/quarterly-connectivity-updat...

Fibre is owned by infrastructure companies that charge standard rates to ISP’s, and the ISP’s compete to provide Internet to consumers.

I've worked on various broadband-related projects as a grant-writing consultant (https://seliger.com/2009/11/13/adventures-in-the-broadband/), SpaceX will definitely not be the worst of the bunch and will almost certainly be among the top quarter.

In my country they simply force the main ISP to open their network on fair rates. No need for and entire new physical network but there is competition.

Reply on news.ycombinator.com