« Home | Quick points on Schiavo case » | Governor Warner, does he really know what "mainstr... » | Two years in Iraq » | March Madness » | A sad anniversary » | Haven't heard much recently from the anti-war Demo... » | The media is reactionary even in Italy. » | Democracy in the Middle East » | StudCo is all over, I'll move on to focusing on bi... » | I should've remembered that StudCo elections alway... »

March 23, 2005

Were all the lives worth it in Iraq? The numbers say maybe.

Having taken some statistics, I have always remained skeptical of any information posed by the media. Polls have proven unreliable (just ask John Kerry). While global warming is the latest environmentalist craze right now, just a couple decades ago, it was global cooling that was the big fear. And now, this.

A number of people have stated that the war in Iraq is not worth it, that the number of people dead is just too high. Many estimates go into the tens of thousands, and some even to 100,000. One such estimate in particular was given by a study from Johns Hopkins University, and was reported by a number of news outlets, including AlterNet. This terrible news would likely turn everyone off to the War in Iraq and make the new democracy seem to have come at too high a cost.

Until you read this article. It also addresses this study. It confirms that the study found a mean estimated value of 98,000 dead Iraqi citizens. However, it points out part of the study that was left out of most reports.

We estimate there were 98,000 extra deaths (95% CI 8000-194 000) during the post-war period.

Now, I hardly expect that everyone knows exactly what this means, so I'll do a quick interpretation of the data. 95% means that there is a 95% chance that the number of deaths falls in the range following it. That range is 8000-194000 people. While this follows a normal curve, meaning that the chance of either 8000 or 194000 is far lower than being 98000, the likely value is far from 98000. Why is this you say? The standard deviation is extremely large (approximately 45000 people). Normally, the standard deviation should be close to the square root of the mean, suggesting a good standard deviation value would be approximately 315. Even giving a little (lot of, really) leniency, this is far from a value of 45000.

Just wait, it gets worse. While it may have been unintended, there were biases that shifted the numbers. Those conducting the polls were limited in where they could go. They also went to areas that generally were more populated; these areas of course were more likely to be hit because they would likely hold terrorists and insurgents. And when you consider that this poll did not include Fallujah (an area considered to be extremely hard hit), it could've been more inaccurate.

Interestingly enough, these numbers lead to a mortality rate of 7.9 per 1000. This might seem high, except most estimates place the mortality rates from 1985-2003 at around 6.5-8.1 people as well, showing that post-war Iraq has seen about the same level of danger as it had before then.

This study is not mathematically sound, and must be completely discounted. While it is extremely likely that the number of deaths is greater than 8000 as the study suggests, the number of deaths more likely falls lower than 90,000 people. Chris Suellentrop, the writer of this article, compares this study to others and comes to a more reliable range of 15,000-30,000 (unlike the JHU study also including during the war).

Note: I am not disparaging those who conducted the JHU study. They had a difficult job and did the best they could. They may have been a little irresponsible for not emphasizing the huge variation, but the bulk of the blame lies on the MSM (main-stream media; obviously, I don't take that literally, and get used to this acronoym, it is commonly accepted and I will likely use it often). The variation was clearly stated in the report, and it was "conveniently" left out by the MSM. For this reason, I push anyone reading this blog to remain skeptical of what they hear on the evening news; they may be well intentioned in what they are trying to communicate, but the information can often be inaccurate and sometimes even deceptive.