In the heated debate over U.S. Internet policy, South Korea is often held up as a model of how the Internet should work. As of Q2 2014, South Korea has the fastest average Internet speed in the world at 24.6 Mbps -- more than double the U.S. figure of 11.6 Mbps. Internet connections are also significantly cheaper than in the U.S.
It's fashionable to blame this on low competition in the U.S. broadband industry and corporate greed at top broadband providers like Comcast (NASDAQ:CMCSA) and Time Warner Cable (UNKNOWN:TWC.DL). However, the truth is a little more complicated.
It comes down to population density
There are obviously multiple factors behind the difference in Internet connection price and performance between the U.S. and South Korea. Population density is by far the most important factor, though.
South Korea is one of the most densely populated countries in the world. It has a population of 50 million crammed into less than 40,000 square miles. (If it were a U.S. state, it would be first in population but 38th in land area.) Its population density of roughly 1,300 people per square mile is more than 15 times higher than that of the U.S.
In a study of South Korea's Internet infrastructure, the International Telecommunications Union cited "unique housing patterns and high population density" as key factors behind its world-class broadband speeds. More than 60% of the population lives in apartments, and most people live within four kilometers of the local telephone office.
This is very significant because DSL on traditional copper wires only provides fast Internet if you are located near the local switching station. Thus, South Korea's high population density -- and the prevalence of apartment living -- makes it very cost-effective to connect people to a speedy Internet infrastructure.
It's no surprise that many of the other countries and territories that have significantly faster average connection speeds than the U.S. have a much higher population density. Examples include Japan, Hong Kong, the Netherlands, and Switzerland. Other sparsely populated countries like Canada and Australia have slower average broadband speeds than the U.S.
The cost of suburban sprawl
Thus, if you want to know the real source of high broadband costs and relatively low speeds, look no further than U.S. suburban sprawl. Living in a detached house is typically considered a key part of the "American dream," but this has a very real cost in terms of Internet connectivity.
With most Americans living too far from the local telephone switching station to get fast DSL on copper wires, U.S. telecom firms have needed to invest heavily to enable faster connection speeds.
This has included ambitious fiber-to-the-home deployments in limited service areas. This involves replacing traditional copper telecom wires linking each subscriber address to the central switching station with new fiber-optic cables that can support fast data transfer rates. However, fiber-to-the-home installations can cost $2,500-,$5,000 per subscriber.
AT&T (NYSE:T) has gone a different route to compete with cable Internet providers. Rather than investing exorbitant sums in fiber-to-the-home, it has adopted a compromise solution called fiber-to-the-node.
This involves building fiber-optic networks into residential neighborhoods, but not all the way to customers' homes. AT&T claims that the nodes (where fiber switches to copper) are within 2,000-3,000 feet of customers' homes, on average. This strategy reduces capital costs to $300-$500 per subscriber, according to industry sources.
All in all, the popular conception of cable and telecom companies as greedy penny-pinchers is overstated. The U.S. broadband industry has spent $1.3 trillion on capital expenditures since 1996, according to the United States Telecom Association, a telecom industry trade association. Of that total, 69% (about $900 billion) went into fixed broadband lines, with the remainder spent on wireless connectivity.
There are about 90 million broadband households in the U.S. today. Total wired broadband investment in the U.S. since 1996 thus works out to about $10,000 per household. By contrast, public and private entities in South Korea invested a total of $32.5 billion on broadband in the decade from 1995-2005.
Even adjusting for population, this is significantly less than what U.S. cable and telecom firms invested. No wonder why broadband is more expensive here!
Is competition the problem?
Some people argue that a lack of effective competition is the real reason why U.S. residents pay a lot for Internet connections that are not even close to being the fastest in the world. Telecom firms are natural monopolies -- it would be very costly for a potential competitor to build parallel connections in order to offer a choice of telecom providers at every address.
In South Korea and many European countries, government regulators have forced the telecom monopolies to sell access to their subscribers on a wholesale basis. This allows numerous Internet service providers, or ISPs, to compete for customers with different prices and service offerings without each having to spend billions of dollars to build duplicate connections to each subscriber.
However, this strategy doesn't always lead to superior broadband performance. Over the past decade or so, the U.K. has been progressively introducing competition between ISPs. It forced British Telecom and Kingston -- the country's two telecom giants -- to allow competitors access to their last-mile customer connections.
Nevertheless, Internet is no faster in the U.K. than it is in the U.S. As of Q2 2014, the average connection speed in the U.K. was 3% slower than in the U.S., while the average peak connection speed was 3% higher than in the U.S. This is true even though Britain's population density is eight times that of the U.S., which should make it easier to provide fast Internet service.
Where's the money?
Critics of the U.S. Internet system argue that with more competition, the U.S. could have cheap, speedy Internet like South Korea. However, they can't explain where the money will come from to fund faster speeds with lower costs to consumers.
If the big U.S. ISPs were earning huge monopoly profits, competition might be an easy solution. However, even at the biggest and most profitable broadband providers, operating margins are around 20% and after-tax profit margins are in the high single-digits or low teens -- relatively reasonable levels considering the capital-intensive nature of the business.
Even if these companies are forced to open their last mile connections to competition, they would still have to be able to cover their costs and earn a margin. (The competing ISPs would also have to be able to earn a profit.) This would prevent major price cuts or major increases in investment.
More competition in the broadband industry might help improve the industry's terrible customer service standards. It also would give consumers more service tiers to choose from. However, the two things people care about most -- price and speed -- wouldn't change much. Suburban sprawl means that the U.S. won't be able to match South Korea's cheap, speedy Internet service.