WEBVTT 00:00:00.000 --> 00:00:04.160 When we talk about the monthly cost of home internet, we usually focus on how 00:00:04.160 --> 00:00:10.080 annoyingly expensive it is. But another odd aspect about internet prices is how much they can vary 00:00:10.080 --> 00:00:16.080 from place to place, even inside the same country. It's generally weird that you need to pay around 00:00:16.080 --> 00:00:21.280 $100 a month for internet service in Minneapolis that is roughly the same quality as internet service 00:00:21.280 --> 00:00:27.280 you can get for just $60 a month in Fort Worth, Texas. The reason most commonly cited for 00:00:27.360 --> 00:00:32.640 high internet prices is lack of competition between internet service providers or ISPs. 00:00:33.200 --> 00:00:39.200 And that's not wrong, but it's worth unpacking why that is. And no, the answer isn't just 00:00:39.200 --> 00:00:44.560 because they're dicks. They are, but that's not the whole story. The basic problem with the 00:00:44.560 --> 00:00:49.520 free market approach to broadband internet service is that it's actually incredibly hard for new 00:00:49.520 --> 00:00:55.200 competitors to join the market and potentially financially ruinous even to try. Similar to 00:00:55.200 --> 00:01:00.720 water pipes, electrical wires, and rail lines, internet cables are both expensive to install and 00:01:00.720 --> 00:01:06.560 basically useless until you create an extensive network of them. This creates an extremely high 00:01:06.560 --> 00:01:12.560 upfront cost that a private company will typically only take on if there's a near absolute guarantee 00:01:12.560 --> 00:01:17.680 that they'll be able to recoup their investment and more. This often results in what economists call 00:01:17.680 --> 00:01:23.200 a natural monopoly. These are usually relatively small markets where the barrier to entry and 00:01:23.200 --> 00:01:28.160 startup costs are so high that once a single service provider establishes itself, no other 00:01:28.160 --> 00:01:34.080 competitor will even bother to try. Though in larger markets, this is more likely to result in 00:01:34.080 --> 00:01:39.760 a duopoly or oligopoly with a small number of large stable players who mostly stay out of each 00:01:39.760 --> 00:01:45.520 other's way. You might think of the entire United States as a pretty big market, but because each 00:01:45.520 --> 00:01:50.880 company needs to lay infrastructure into every single little community, each contiguous urban 00:01:50.880 --> 00:01:55.680 area is effectively its own isolated market. Therefore, many of us only have a choice between 00:01:55.680 --> 00:02:00.720 one or two ISPs, pretty much for the same reason that everyone else has only one choice for their 00:02:00.720 --> 00:02:06.240 water utility. There's only one set of pipes leading to the house. What determines whether a 00:02:06.240 --> 00:02:12.480 particular region winds up with one, two, or three ISPs is a combination of geography, population 00:02:12.480 --> 00:02:18.720 density, and local regulations. The harder it is to install and replace cables and other infrastructure, 00:02:18.720 --> 00:02:24.400 the more customers you need to absorb the fixed cost of installation. This can be a big problem 00:02:24.400 --> 00:02:30.160 for countries with large, sparsely populated rural areas like Canada, the United States, 00:02:30.160 --> 00:02:35.280 and Australia, but it's also a problem for places with lots of rivers, deserts, and mountains 00:02:35.920 --> 00:02:41.680 like Canada, the United States, and Australia. It's exactly these kinds of high infrastructure 00:02:41.680 --> 00:02:46.080 costs that could make satellite internet a potential game changer for many rural markets 00:02:46.080 --> 00:02:51.040 if it became significantly cheaper in the near future. In the meantime, however, shooting 00:02:51.040 --> 00:02:56.320 satellites into space is still relatively expensive. For now, the places globally where 00:02:56.320 --> 00:03:02.400 broadband is both extremely fast and relatively cheap typically have extremely dense urban populations 00:03:02.400 --> 00:03:08.800 and little or no rural population such as Hong Kong and Singapore must be nice. Of course, 00:03:08.800 --> 00:03:13.200 just because Monopoly is natural does not mean there's no funny business involved, 00:03:13.280 --> 00:03:18.560 and it doesn't mean there's no way to fix it. The problem with natural monopolies is that 00:03:18.560 --> 00:03:23.600 they make the free market not act like a free market. They discourage rather than encourage 00:03:23.600 --> 00:03:28.960 competition because when actually competing is expensive and hard, it's much easier to succeed 00:03:28.960 --> 00:03:35.840 through lobbying and bullying small competitors. In order for a small ISP to grow, it will need 00:03:35.840 --> 00:03:40.640 to get any number of government permits, all of which can be tweaked to make them harder than 00:03:40.640 --> 00:03:45.600 necessary to acquire. Plus, the small company will typically need to be able to build enough 00:03:45.600 --> 00:03:51.920 infrastructure to actually appeal to some customers. In the meantime, they might need to rent access 00:03:51.920 --> 00:03:56.240 to some of their larger competitors' infrastructure and bandwidth, something which those competitors 00:03:56.800 --> 00:04:02.080 might not be willing to do unless compelled by the law. Most countries do require this kind of 00:04:02.080 --> 00:04:06.720 cooperation for just this reason. Theoretically, these small companies could eventually hit a 00:04:06.720 --> 00:04:12.080 tipping point and compete completely independently, but long, long before they hit that point, 00:04:12.080 --> 00:04:16.800 they'll hit a different tipping point where the smartest thing to do is to sell themselves 00:04:17.440 --> 00:04:22.720 to the biggest competitor. In a natural monopoly scenario, the incentive of the primary players 00:04:22.720 --> 00:04:28.240 is to gradually gobble up all small companies and eventually merge into one massive company 00:04:28.240 --> 00:04:33.520 so they can make bigger margins and stop investing so much in infrastructure. In this situation, 00:04:33.520 --> 00:04:39.120 antitrust action and legislative prevention of further mergers are likely the only option to 00:04:39.120 --> 00:04:45.440 restore competition. Regulation is messy, and that's why it makes everybody either bored or angry. 00:04:45.440 --> 00:04:51.280 Here again, Canada is a good example. We have both ineffective antitrust regulators and 00:04:51.280 --> 00:04:57.040 high regulatory barriers to entry. While most of the country has abominable internet prices due 00:04:57.040 --> 00:05:03.120 to astronomical infrastructure costs, Ontario and Quebec, the most populous provinces, have 00:05:03.120 --> 00:05:08.160 worse prices than the next two most populous provinces, British Columbia and Alberta, due 00:05:08.160 --> 00:05:16.480 to greater consolidation among ISPs. So is the answer here even more government involvement? 00:05:17.040 --> 00:05:22.720 Some telecom industries are fully nationalized and run either directly by the government or by a 00:05:22.720 --> 00:05:28.560 semi-independent public corporation, while other countries heavily subsidize building infrastructure 00:05:28.560 --> 00:05:34.080 as a way of getting past the initial barrier of upfront investment. While this probably doesn't 00:05:34.080 --> 00:05:38.800 reduce the overall cost of the infrastructure, it can at least encourage faster development 00:05:38.800 --> 00:05:44.960 by spreading out risk across the entire tax base of the populous, rather than concentrating it into 00:05:44.960 --> 00:05:50.880 a single company with a relatively small pool of customers. An interesting American example is that 00:05:50.880 --> 00:05:55.840 of Chattanooga, Tennessee, which is very fun to say, and has some of the cheapest and fastest 00:05:55.840 --> 00:06:00.800 internet speeds in the United States. And it was the first US community to adopt Gigabyte speed 00:06:00.800 --> 00:06:06.400 internet back in 2010. Chattanooga, I'll never get sick of that, businesses can even get 25 Gigabyte 00:06:06.400 --> 00:06:12.480 internet as of 2022, though it is understandably pricey. In this case, the Chattanooga Electric 00:06:12.480 --> 00:06:18.000 Power Board is the city-owned ISP and operates essentially like a normal local utility company. 00:06:18.000 --> 00:06:22.240 Some governments even take the middle-of-the-road approach of creating a public corporation 00:06:22.240 --> 00:06:27.440 without a monopoly so that it can provide a consistent competitor to private companies 00:06:27.440 --> 00:06:32.800 in the same market, such as SaskTel, the public corporation created by the Canadian province 00:06:32.800 --> 00:06:38.480 of Saskatchewan. You've heard of it. For context, this is the very rough Canadian equivalent of 00:06:38.480 --> 00:06:44.240 Montana or one of the Dakotas. Despite its large area and low population density, many Saskatchewan 00:06:44.240 --> 00:06:49.600 residents have access to cheaper, faster internet than what's available in large dense population 00:06:49.600 --> 00:06:56.080 centers like Vancouver. But if you hate big government exactly as much as you hate big 00:06:56.080 --> 00:07:01.840 corporations, there is always the option of a telecommunications cooperative where the ISP 00:07:01.840 --> 00:07:07.520 is collectively owned by the customers. While they can vary a lot in size, some are quite small. 00:07:07.520 --> 00:07:12.480 Telecom co-ops are found throughout the United States and around the world, often in rural 00:07:12.480 --> 00:07:17.680 areas that are typically neglected by traditional telecom companies and ISPs. Not only can this 00:07:17.680 --> 00:07:21.920 be a way of providing cheaper, higher-quality telephone and internet services in smaller 00:07:21.920 --> 00:07:27.680 markets, it can sometimes be the difference between decent internet access and no access at all. 00:07:27.680 --> 00:07:31.760 No matter what we wind up doing, this problem of varying prices across regions 00:07:32.320 --> 00:07:36.480 is not something that's going to get better on its own. It's going to require concerted effort, 00:07:36.480 --> 00:07:42.160 cooperation and investment to change the incentives that led us here. But that change 00:07:42.160 --> 00:07:46.240 is possible. Thanks for watching guys. If you liked this video, why not learn more about how 00:07:46.240 --> 00:07:52.640 the internet works by checking out our video on top-level domains? Surprisingly, very interesting.