Thursday, 30 July 2009

Everything Fail

Whilst I was gallivanting around Colchester Kevin Drum churned out a pair of very interesting posts. The two are almost entirely unrelated, beyond a shared theme of maths, assuming utilitarian theory counts of maths (or averages, for that matter, which I would assume comes under basic numeracy, which neither I nor my proudly mathematically incapable friends should come under the heading of mathematics). I'll get to the more complex one later on (partially to give myself time to decide whether or not I want to pick a side), but this was so hilarious I couldn't leave it alone. Ever wondered why Canada has a higher life expectancy than the States? Mr Bill O'Reilly has the answer!
[W]e have ten times as many people as [Canada does]. That translates to ten times as many accidents, crimes, down the line.
Bear in mind that this was in response to a letter. O'Reilly had the chance to consider his reply. Drum describes this as flubbing fourth grade arithmetic, but I don't think that's the problem. Knowing how to calculate an average might be fourth grade (that sounds a little bit early, actually, but I don't want to bet my more recent experience with the English maths syllabus trumps Kevin's distant memories of an American one); O'Reilly's mistake is one of basic English. He simply cannot understand what a life-expectancy is. I literally can't find any way to understand the mistake (and I taught this subject to eleven-year olds for three years, so I've seen my share of misconceptions). For a moment I assumed he believed life expectancy is the sum total of everybody's predicted age, but that doesn't make sense because then the States would be higher (it would also put the number in the billions, which might have made even O'Reilly think). So he must know that the life expectancy is how old each individual person is expected to survive, but apparently can still think that adding more people into the pool will automatically drag the number down. Of course even by his insane logic, more people means more success stories and medical miracles, so the whole thing should be a wash, but the fact O'Reilly didn't mention that can at least be written off as deliberate wingnut BS.

I've just realised that I'm attributing outrageous mendacity to O'Reilly in order to give him credit. Remind me again why this gomer is right-wing royalty?


Senior Spielbergo said...

Is there a link to the full transcript somewhere? I can only find video and I have no sound at work :¬(

I’m guessing there was something more to his answer as it doesn’t seem to make any sense stand alone and especially as a response to a question on life expectancy. I would expect there to be a follow up sentence or two seeking to explain why this mean your life expectancy is lower. I’m not seeing how it’s going to result in a sensible response either way (although I’m forced to admit that as a standalone statement it is technical logical), but I’m curious to see how he attempted to dig himself out of that hole or if he just didn’t bother.

SpaceSquid said...

"Is there a link to the full transcript somewhere?"

I can't find a transcript (FOX doesn't provide free transcripts, apparently), but his tone of voice, coupled with the fast-paced nature of the segment (where he reads out bits from letters and comments on them) imply that he was finished. What I gave you seems to be eat (barring him reading the letter itself).

"although I’m forced to admit that as a standalone statement it is technical logical"

Presumably by "standalone statement" you mean ten times as many people means ten times as many accidents and crimes. Frankly even that isn't logical. It would be logical to conclude that more people in the same society will lead to more accidents and crimes in that society, but comparing across cultural divides requires more thought, and suggesting the two increases are perfectly correlated is somewhat dubious as well (though as a rule of thumb I wouldn't have a huge problem with it).

Regardless, I'm curious as to what exactly O'Reilly could say to get himself out of this. From where he was at the end of the clip he would literally need some way of explaining how multiplying a population can by itself alter a statistic which cancels out multiples.

Senior Spielbergo said...

Basically I was just saying that as a stand alone statement it contained a logical thought progression. I.e it would be a reasonable expectation that if you increase the number of people by 10 you would increase the amount of accidents and the amount of crime by a roughly similar amount (ignoring many other factors of course). It makes no sense as an answer to the question of course.

I more reasonable answer, which would presumably be along the same lines as to what I *guess* he was going for was to highlight the much increased crime related death rate, and the much increased accidental death rate (largely as a result of RTC’s) America has compared to Canada. As this kills off a whole load of young people it therefore skews the life expectancy statistic somewhat. It would be interesting to see what the statistics are for life expectancy purely as a result of death by “natural causes” as that would be a better source of comparison for comparing health care systems (which I’m guessing is what the debate is about).

Senior Spielbergo said...

Googling I have found one set of Data that attempts to do the comparison I have suggested above:

(table little bit down the page)

Not really sure of the validity of the data, as the study it links to does not seem that great, but best I’ve been able to find so far…

SpaceSquid said...

"I more reasonable answer, which would presumably be along the same lines as to what I *guess* he was going for was to highlight the much increased crime related death rate, and the much increased accidental death rate (largely as a result of RTC’s) America has compared to Canada."

That was what I figured would have been "better", especially since it would allow O'Reilly to have a pop at Obama being soft on crime.

Bighead also suggested that you could make a fast and loose argument that America's greater population means a greater population density, which will increase accidents and crime as well. That argument is horribly simplistic, and would need some research, but it at least isn't immediately mathematically wrong. In fact, off the top of my head it seems like a reasonable hypothesis.

I'm not convinced about that study you link to, either, but the chart you mentioned is interesting. Of course, it's important to note that people without health insurance are probably the exact same people more likely to die as a result of crime or various forms of accidents (inadequate housing, workplace safety issues, drug abuse), so cutting them out of consideration may help out America's statistics more than those of other countries.

Pause said...

The only way I can possibly make sense of such a stupid answer is if "crime" was an implicit acknowledgement that gun crime in the US is higher, and significantly more people get shot young.

For someone in such a public position to make the link that high gun ownership rates leads to a high volume of fatal shootings is such a brave... hah hah hah, yeah, never mind.