I've heard about this research study, but I didn't pay that much attention to it until I saw that they had a video of this effect that was rather neat. It also confirms what I had suspected for a while.
The study was done by a group of Japanese scientists/mathematicians about traffic jams that happened without any bottleneck. In other words, there's no obvious obstacles, such as an on-ramp, or a constrictions, etc. that would be obvious causes for a traffic jam. All they did was increase the traffic density, and at some point, there's some "critical density" in which traffic jams simply occurs because people just don't all drive at the same speed.
What is need is that you can actually this taking place in the video that they have included. I've always suspected this. I drive roughly 31 miles each way to work every day. I sometime get stuck in a couple of slow spots where traffic either slows down or stopped for periods of time. Yet, as you you drive some more, you speed back up again as if nothing has happened, and you don't see any reason why the traffic slowed down. I tend to blame it on slow cars in the left lane, but I had no proof that was the usual cause. Now, I have some evidence to back my haunch! :)
 Y. Sugiyama et al., New Journal of Physics v.10 p.033001 (2008).