A decade ago today, on August 14, 2003, the lights went out in the northeastern U.S. and much of eastern Canada, plunging 55 million people into darkness–or at least brownness. The cause of the blackout, the second-largest power failure in world history (following a 1999 blackout in Brazil), was a small glitch in one computer running at one power company’s office in Ohio. This small bug, known in computer lingo as a “race condition,” ultimately cascaded into a disaster that shut down the largest cities in the United States and Canada.

Exactly how the computer glitch led to a massive blackout is extremely complex. The key dynamic of modern power infrastructure is what they call “load”–how much electricity is traveling over power lines at any given instant. Too much load will cause power lines to sag and other disastrous effects. Essentially, power managers are kind of like air traffic controllers, diverting power this way and that, preventing the power equivalent of runway stack-ups. Obviously, if they don’t have accurate information about the state of their system, things can go wrong.

What I find interesting about the great blackout of 2003 is how different it was from the last great power outage to strike New York City–that one occurring in 1977. That famous blackout (incidentally it’s depicted in the movie “Summer of Sam,” which I recently profiled) resulted in widespread looting and anarchy in many parts of the NYC area, with people smashing windows and making off with electronics and other goods before the police could stop them. Like in 2003, the 1977 blackout occurred during the summer, and on a particularly hot day. Unlike 2003, the 1977 blackout began at night. Ten years ago the power went out at 4:13 PM. It was still off for most of the night in many places, but the initial failure occurred during the daytime. I wonder if that was a factor.

But in 2003 there wasn’t much anarchy. Quite the contrary. New Yorkers left their offices on foot, and many migrated to bars and restaurants, who were giving away their perishable foods–especially seafood and ice cream, which would not last very long. Block parties broke out in many parts of the city. (That also happened in 1977). Of course, people had to be pried out of elevators, led up from darkened subway tunnels, and a blackout of any duration is problematic for hospitals and other crucial infrastructure. But you just didn’t see a lot of anarchy.

blackout 2

New Yorkers on foot during the blackout of August 14, 2003.

It was the same in other places. Toronto, the largest city in Canada affected, saw hassle but no violence. Most of the effects related to transportation failures. I’m not sure how likely it is that Canadians would smash windows and make off with TV sets in any condition, but it certainly didn’t happen in 2003.

Why? Was it just the time of day? It can’t be that simple. There must be social factors at work. New York had more police on the force in 2003 than in 1977, but that doesn’t seem to answer it either. Mob psychology is notoriously difficult to figure out. Are we Americans “calmer” or “less prone” to riot today than we were in the 1970s? I don’t know how you’d go about making an argument like that, but there must be something at work there.

I find blackout lore to be very fascinating. Power failures are the catastrophic disaster-but-not-really-disasters of our time. How do we behave during them? Why do we do the things we do? All of these are very interesting questions. Sometime in the future there will be another major blackout. Perhaps conditions then will be very different. While I’m glad the 2003 failure went as “well” as it did, we may not be so lucky the next time around. Time–and darkness–will tell.

Top photo by Brendan Loy. Bottom photo by Glitch010101. Used with attribution CC license and relicensed under same.