U.S. Troops to Iraqis: Leave Your Lights on at Night

A fascinating U.C.L.A. study challenges the success of the U.S. troop surge using a creative metric: satellite images capturing the amount of light at night in Iraqi neighborhoods. The idea is simple: lights at night are an indication of activity. According to the article, no lights imply that people have fled or have been ethnically cleansed. (It also might mean that there is no electricity, or that strict curfews are being enforced, but the authors say that is not a full explanation.)

From the press release:

The night-light signature in four other large Iraqi cities — Kirkuk, Mosul, Tikrit, and Karbala — held steady or increased between the spring of 2006 and the winter of 2007, the U.C.L.A. team found. None of these cities were targets of the surge.

Baghdad’s decreases were centered in the southwestern Sunni strongholds of East and West Rashid, where the light signature dropped 57 percent and 80 percent, respectively, during the same period.

By contrast, the night-light signature in the notoriously impoverished, Shiite-dominated Sadr City remained constant, as it did in the American-dominated Green Zone. Light actually increased in Shiite-dominated New Baghdad, the researchers found.

Until just before the surge, the night-light signature of Baghdad had been steadily increasing overall.

Here’s the question: Will it be enough for the U.S. troops to command the Iraqis to leave their lights on so that things look good for the surge, or will we need to assign troops to shine lights skyward themselves?

(Hat tip: Claude Eilers)


Mabye the lights are off because people now have jobs and have to get some sleep at night?


an increase in combat troops hunting down perpetrators of terrorist acts causing a subsequent decrease in such acts? come on that's preposterous. next you'll tell me fluctuations in solar output might influence the temperature on earth


@#14 J

You know, it's an oft-repeated Colbert joke, but the idea that reality has a liberal bias actually has a grain of truth to it. People who identify themselves as conservative tended to believe falsehoods, *more so* after said falsehoods were later refuted (to control for the bias against perceived liberal media bias, the study replaced the names of media outlets that refuted the falsehoods with Fox News and still got the same results).

Also, you do realize that you are reading the Freakonomics blog - the whole idea is to use unconventional metrics to quantify things. So yes, it may seem comical to you.

Finally, I don't think any reasonable person suggests that the surge was a failure. What reasonable people do point out, is that the surge was but a small factor in lowering the violence, compared to the Sunni Awakening and some strategic shifts unrelated to the troops surge.



What really grinds my gear (thanks peter) is the fact that most people can agree that something has been a success over the past couple years, from the time the surge started to now, and over that time Iraq has turned into a generally safer place. Less soldiers are dying, less civilians are dying (and let's remember most of those are not US-caused causalities, despite what others would have you believe).

Yet we constantly argue about what caused it? To Joe America it really doesn't freaking matter. The people I see arguing it have no need to know this, no actual intel, and just are arguing talking points or data manipulations of major affiliations usually ignoring any hard data.

The simple matter is to the public the message should be simple. Iraq is a safer place. To the policy makers (and those involved with understand the situation) should examine why it's working but I don't see the necessity to air "this is why it's working" in the public view, especially when the explanation and documentation of the work is poorly done.

Why can't we just celebrate something is going right, and let's get the people who should focus on what's going right to do that, rather than focus on telling us what's going right every five minutes and then defending it ad nauseum?



Simple is the correct description of the idea. There are so many holes in this I don't know where to begin.

Raj Pandravada

Studying night-light signatures seems to be a clever way to detect activity, or inactivity. But it can also be misleading, IMHO.

Here's a link to a 'Strange Map' of the Korean peninsula.


According to the study, most of North Korea must be unpopulated or extremely inactive. This is of course untrue - there are 27 million people in the 'dark' half of Korea....

Even so, kudos on a novel way to try to debunk this 'surge is working' assertion that's been floating around.


A few weeks before one soldier was placed in Bagdad for the "surge", field reporters for NPR were reporting a reduction in violence due to the fact that most of the neighborhoods had already been ethnically cleansed. The reporters also attributed the reduction in violence to paying the Sunni insurgents and the Shiite cease fire that took place the summer before. These reports are archived and be confirmed by anyone who doubts this truth. If you want to argue that the surge worked, I would suggest that you back it up with facts.


Shouldn't the title of this post be "Academic Leftists Go To Comical Lengths To Deny Success In Iraq". Or maybe just "Alan Sokal Is At It Again". Has this been published in any social science journals yet?

"It would also seem that the number of Mercedes dealerships opening in surge-targeted neighborhoods has been much smaller than one would expect"

Other more convincing metrics they could have used: Baghdad lacks even a single Starbucks outlet, and there's no IKEA either. Clearly the surge has been a complete failure.


Perhaps the authors are confused, thinking that "surge" refers a burst of electricity.



As Einstein said:

"Everything should be made as simple as possible, but not simpler."

This is simpler - and bad. And bad is not good. The rest of your post suffers from this problem.

A credible test? Sure but only of how much light you can measure...everything after that is just WAG and storytelling.


Aside from the aforementioned problems with the study, residents also leave lights off at night so as not to provide a target for insurgent mortar attacks. So, although people living in those neighborhoods are increasingly feeling safer, they may still be playing it as careful as possible.

Nonetheless, I think this is a useful tool for understanding success in the overall rebuilding effort; not only in Iraq but in any post-conflict society. Obviously, used alone it has a lot of problems, but in conjunction with other measures is a helpful indicator of overall security and development in a given area.


@ #16 holycalamity

Please learn a little bit more about the surge before making such a misinformed claim as "...and some strategic shifts unrelated to the troops surge." The "surge" is shorthand for the shift to the counterinsurgency doctrine--which had recently been codified in Army field manual FM 3-24 [warning: PDF at link], perhaps not so coincidentally co-authored by then Lt. Gen. Petraeus--that was supported by an increase in troop strength. Hardly unrelated.

And how about trying to think in terms of a system rather than a single linear equation. Do you think it at all possible that, for instance, the new COIN doctrine helped the Anbar Awakening, which had already begun, to strengthen and spread, and which, in turn, allowed the USMC and U.S. Army to continue to push out into neighborhoods rather than pull back to forward operating bases?



Maybe the lights are off because people no longer have to stay up at night and guard their homes. If I were afraid of people attacking me at night, I'd stay up with the lights on.

Paul K

Actually, as even the commander in charge has noted (in complete opposition to McCain and the other "surge is the answer" blowhards), the real change has been a combination of the Sunni Awakening (making them our allies vs. enemies), the improved humint (intelligence from people) allowing better targeting, and finally getting the power restored to a level that people feel better about their situation. The so-called surge has played a minor role in that, although was started at the same time.


@#6 BT

If people have fled in large numbers or were ethnically cleansed (did we stop using the perfectly cromulent "holocausticized"?), then naturally, wouldn't we see drop in reported fatalities and car bombs (what's the point of bombing a deserted town?) as well?

In any case, this study notwithstanding, let's not give too much credit to the surge without acknowledging the Sunni Awakening, which happened before the surge and is most likely more responsible for the decline in violence.

Robert L. www.neolibertarian.com

So before the surge all the liberals said that the surge would either make things worse or do nothing at all and now we have yet another think piece explaining that the decline in violence just happened to coincide with the surge and was actually due to _________ (fill in the blank with anything else).

Plus, if this study is correct about de facto ethnic cleansing being completed before the surge, violence levels would have already dropped before the surge and then stayed level throughout the surge. IIRC that is not what happened at all.

I've hit a point where I just assume any study like this is purely politically motivated and even a cursory look at the study seems to bear that out.


Not necessary, really -- Barack Obama has already declared the the surge has "succeeded beyond our wildest dreams."

What higher authority do we need?


The best indication that this is a flawed study was the comment by one of the authors:

""We had no axe to grind"

In academic circles, this is usually a very strong indication that yes, they were furiously grinding away.


Simple is not bad, Charles. Simple is good.

The question is whether the drop is violence is a result of fewer people in the neighborhood, or actually due to improvement in the rule of law.

The data is clear. There is less nighttime light. There are controls -- looking in other neighborhoods and other cities.

So, what inference can be make from this. They offer at least one. Come up with your alternative explanations/inferences.

Perhaps a power outlet on that one night, suggested Mr. Durst. That is relative easy to verify, right? Did this neighborhood suffer more power outages at that time than the control neighborhoods and cities, right?

The fact is that many sources on the ground have been reporting ethnic homogenization of Bagdad neighborhoods for years, now. Here is some data that supports that. (In other words, the theory makes a prediction that can be verified or shown false. In this case, it is that there would be less nighttime light in those neighborhoods.)

That's a credible test of a valid theory.



Number of bomb attacks or human fatalities is not good enough, is it?

In a non-freakonomic sense, I see too many flaws in this study upon first glance.