Study shows U.S. broadband speeds continue to lag

The average download and upload speeds for broadband services across the U.S. have remained relatively unchanged over the past year as the U.S. continues to lag behind other countries in terms of broadband speeds, according to a report published by the Communications Workers of America labor union.

The report, released Tuesday, consists of data from 230,000 online speed tests across the U.S. In its Speed Matters blog announcing the study's results, the group highlighted that the average download speeds increased only 0.4 megabits per second to 2.3Mbps. By contrast, the average download speed in Japan is 63 Mbps and in South Korea the average download speed is 49 Mbps.

As for upload speeds, the average in the U.S. was only 435 kilobits per second.

The CWA is using the report to lobby lawmakers to develop a national broadband policy. Specifically, the union is backing a Senate bill called the Broadband Data Improvement Act (S.1492) that calls for the government to define advanced telecommunications services so that the Federal Communications Commission can more accurately define broadband conditions in the U.S.

The union's president, Larry Cohen, said that improving broadband service is critical for the U.S. to remain competitive in the global market.

"This isn't about how fast someone can download a full-length movie," he said in the blog. "Speed matters to our economy and our ability to remain competitive in a global marketplace. Rural development, telemedicine and distance learning all rely on truly high-speed, universal networks."

Of course, Cohen's union has a vested interest in whether companies such as AT&T and Verizon Communications upgrade their networks to handle faster broadband speeds. The more than 700,000 members of the CWA union are workers who would help build these networks.

That said, there have been other studies that have shown the U.S. lagging behind other countries in terms of broadband penetration and speeds. Other groups have highlighted this research to also push for a national broadband policy.

[inachu]

Korea is so far advanced and we lag behind all for the mighty dollar to squeeze the most out of the lowest quality we can give to consumers and tell them its gold.

[Killer_KF]

This is what happens when companies monopolize the market

[CmdrRickHunter]

Here's a far more interesting question: what percentage of our bandwidth which we are supposed to have is actually available? The packet shaping of P2P proves beyond a shadow of a doubt that our last miles are over sold, and there isn't really 2.3mpbs per person. In the past that was fair (we didn't have a good way to use the bandwidth). In the modern era, however, with all of the visions mentioned in this article, bandwidth is going to actually have to be sold as it really is - the value of its space on the mighty backbones of the internet. how does Korea compare in that? I don't know.

[CmdrRickHunter]

Annother factor is density. I don't know much about Korea's population, but Japan is VERY dense. The more dense your population, the lower the entrance cost on new technologies. The lower the entrance cost, the sooner technologies are adopted.

[sanenazok]

@Cmdr: you're right about density. It's silly to compare US to Korea and Japan. The closest comparison would be Canada which is at the same level of development, large, and unevenly populated.

There's also a big, big flaw in this report. The U.S. speeds are based on self-reporting while the International data is sourced from something called "Information Technology and Innovation Foundation" which is a (drumroll) think tank (aka lobbyist). The source of the data is significant since it makes the U.S. sample NON-SCIENTIFIC and the international data questionable. Even though the size of the U.S. data is large it's still not adjusted for basic factors, like for example only people who have a speed issue will go to their website to test the speed. This is way surveys have to be scientific not self-reporting.

[tajna_rabota]

No, Canada would not be a good comparitor, as our population is very sparse and spead out over vast distances. Canada has about 33 million people in the second largest land mass in the world. While 90% of us live within about 200 miles of the US boarder, the simple fact remains that this is a small number of people over a huge distance.

I see your point as we are about the same development level and we are 'unevenly populated' - but population density wise, the US is to Canada what Japan is the US.

The US is unique because it has huge cities, high population density (relative to the size), but it also has a 'suburbia' element not found in most other developing countries. Most developing countries in this study have people living in apartments, townhouses, and other multi-family dwellings. I'd wager that a large portion of US broadband customers live in single-family houses on large lots - so the question is whether there is a decent comparitor out there?

[Inetsec]

What about all the "dark fiber" that we have all been hearing about over the years? Light it up already!

[bobanavrin]

I don't know where you get your info , most of the dark fiber is now in use.

[skillingssucks]

The problem is the "last mile"...the connection to residential homes.

[William Crow]

My cable internet hookup, often advertised as super hi-speed, is often about the speed of dialup.
At these slow speeds I notice that advertisements, as opposed to content, is always downloaded first.

[sanenazok]

Welcome to a comcastic connection. Well you might not have Comcast, but if it's a shared connection I'm not surprised that you're not happy with it. At least my DSL is a constant and equal connection of ... 300k! Yey it's the best AT&T can do in the middle of Chicago.

[SactoGuy018]

I think people forget that in Europe, South Korea and Japan the population density is high enough that telco's can afford the exorbitant expense of wiring up everyone for broadband connections.

Because of the sheer size of the USA and the sprawled-out nature of many US cities, the cost of installing broadband is much more expensive here in the USA, and only in very recent years did we finally get the technology to make either ADSL or cable broadband widely available to customers in metropolitan areas.

[chash360]

Dense or sparse we need to upgrade our networks, we need to remain competetive, need communication we can rely upon. This work would create jobs, that we desparately need, ones that could not be offshored to a 3rd world country, for slave labor wages.

That said, there is a much easier way to provide universal ubiquitous coverage without the enormous expense of laying lots of fiber everywhere. A GPS dynamically routed wireless mesh network. If the FCC would just create a device and protocol specification for such a network, device manufacturers would build them, and you would eliminate the need for 'licensed' service providers. When it is totaly wireless, it is the devices that provide the service, not some company you are paying to have access. With short range channel hopping, and frequency reuse, you can efficiently use the EM spectrum and never run out of bandwidth. The critical specification is just that every device must be able to route and transfer more traffic (a multiple of), that which the end user of the device is consuming. Example is I have a device which provides me a 100mbps pipe to use, it should be able to handle and route 1000mbps of ambient local traffic. That way the more devices in use, the more bandwidth there is, we would never outgrow it. The denser the population becomes the more bandwidth is available, and the more access points to the network are present. With physical routing based upon GPS coordinates security is greatly enhanced because it ties your communication to real physical location end points, that can not be faked or forged. Example: Everyone has access to the GPS signals (free of charge), I connect to a website hosted by a store in the local shopping mall, I route to the physical location of that store, to see their website. The resolution of the location address is precise enough that even someone sitting inside the store with a 'hacked' unit is percieved by the network to be physically located in a different spot than the true server, and thus is not able to 'hijack' data. The network itself can refuse connection to a device that reports its location different from where it actually is because it can double check the timing of the radio signals to verify its location is true. A unit attempting to mimic another units location is easily flagged as an imposter.

For all those concerned about big-brother, and privacy, the physical location data does not (and should not) actually have to ever leave the networked device, and does not have to be part of your data. It is only used for establishing routes and connections between devices, and is independent of the data being transferred, or any conversation traveling along those routes. It is a simple matter, and essential, that all of the data transferred be cyphered, using dynamic unpredictable information known only to the devices engaged in the connections. This prevents any easedropping.

The analog TV spectrum was the perfect choice for this type of network, and the selling of it to the highest bidder when it belongs to all of us, is an outrage. Tell the FCC you want your airwaves back, tell them it is their responsibility to efficiently manage the EM Spectrum, not sell it, nor outsource such management to 3rd parties, nor create communication monopolies.

[regulator1956]

"If the FCC would just create a device and protocol specification for such a network"

The FCC doesn't do that. They work with and review what others create and develop.

[chash360]

The FCC does set the power spectrum limits, specifies which bands to use for which kind of services/devices, etc. Others have created such device specs like the one I have proposed but as of yet the FCC refuses to support them by allocating the proper bandwidth as license free. In the ideal situation, since it is the FCC's mandate and charter to manage and allocate EM Spectrum usage, it should be the FCC that defines a protocol for devices to adhere to, for dynamically using and adjusting spectrum usage in a device. In doing so it would create a method for the FCC to actually be able to manage and even reallocate spectrum usage, by publishing updated power spectrum tables to these devices, rather than the huge undertaking that is occuring with the Analog to Digital TV transition. You would not need to get new devices, or even be aware of an allocation change, the devices could simply start using the new frequencies, or stop using old ones, if the device was capable of it. If applied to all future digital wireless devices then the entire spectrum could be easily managed in realtime, not these long drawn out transition periods. No more spectrum auctions, and no more lack of bandwidth available. With dynamic management there is plenty of bandwidth for everyone, it is no longer a limited resource (there is only so much a single user can actually consume at any point, that being the equivalent limit of our sensory input, which is by the way much more than any download speed yet offered anywhere in the world) and it takes only a tiny fraction of the EM Spectrum to actually deliver this much information.

[nickj1088]

I strongly agree with the comments about population density. A direct comparison between the U.S. and a country the size of one of our fifty states is neither fair nor accurate. That said, the U.S. is certainly better than Russia or China or Brazil, countries of similar size (even though they are developing). I think the future in American internet access should come from wireless technologies (think beyond WiMax). To do that, there needs to be a far greater emphasis in the country on math and science. That's how we surged ahead in the first place, and it is why we are starting to fall behind now.

[DrewClark]

At http://BroadbandCensus.com we believe that it's important to have carrier-by-carrier information about broadband availability, competition, speeds and prices. I go into this in my latest post, "CWA Publishes State-by-State Download Speeds. How About Carrier-by-Carrier Speeds?," at http://broadbandcensus.com/blog/?p=416

Drew Clark
Editor and Executive Director
BroadbandCensus.com