by Daniel on January 27, 2006
Here we go again … this story has been making the news recently, about some types at the Max Planck Institute for Dynamics and Self-Organization. They have got some data from a site that tracks dollar bills and used it to make a statistical model of human travel, which will apparently save us all from bird flu.
I saw the news story in the Guardian and thought “I wonder if they will discover that the phenomenon they are studying is described by a power law?” … and the answer is yes they do.
I then thought “I wonder if they will more or less entirely ignore the existing literature on epidemics?” … and as far as I can see the answer is yes they do.
And finally I thought “It would be presumably too much to hope that there is a serious attempt made to test the power law model against alternative distributional assumptions” … and the answer is yes, it would be too much to hope.
They’re also quite naive about banknotes. At one point in the paper, they seem to believe that bank notes “exit the money-tracking system for a long time, for instance in banks”. Of course, banks don’t contain large piles of banknotes sitting around for months on end; the only bank notes you will find in a bank are the ones being stored in the ATMs and the tills, and these obviously turn over pretty rapidly. Furthermore, I would guess that the average human being makes fewer journeys to a banknote sorting centre than the average banknote (I suspect that several of the odd statistical properties that they found relate to the locations of sorting centres in the US). Humph.
People have been laughing at physicists for this sort of thing for quite a while now but with no noticeable effect. It’s all very laudable that they’re taking a bit of interest, but we still clearly haven’t conquered the hard-science arrogance problem yet …
Edit: Actually thinking about it, this is not an entirely worthless paper. If they are correct to claim that epidemic models use a diffusion process to model travelling humans (and there is nothing in the paper to make me think they have done an exhaustive literature search) then epidemic models are being silly and should stop. Human beings don’t diffuse – they go from one specific place to another. Otoh I seem to remember a lot of quite detailed modelling of SARS which certainly did take airline routes into account so I suspect that the diffusion assumption is really only found in toy models in basic-level textbooks.
by Daniel on January 23, 2006
Henry gave me the heads-up on this extraordinary story from Institutional Investor (it is worth watching the interstitial ad in order to read the story as it really is dynamite). It’s an account of the dog’s breakfast that was the Harvard Institute for International Development’s mission to Russia. I learned a few things I didn’t know from it – particularly, it was interesting the extent to which Jeffrey Sachs, who suffered quite a lot of damage to his reputation through being the titular head of HIID at the time it all blew up, wasn’t actually in charge, and to which Andrei Shleifer and his wife were involved in the whole thing to an extent which it is rather implausible to dismiss as the work of a head-in-the-clouds academic who didn’t realise that there might be a conflict of interest. It adds quite a lot to David Warsh’s excellent coverage of the same story, and indeed to the cacophony of fuck-all which has been our beloved mainstream media’s coverage of what is quite visibly one of the most interesting and scandalous tales of the 1990s (I’m linking to Warsh’s own story on the II story because it contains a few quotes that are in the paper version but not the online one, but it is well worth poking around in the Economic Principals archives for other stories on this subject).
[click to continue…]
by Daniel on January 11, 2006
Steven Landsburg, perennial bete noire of people who want to say that economists aren’t an entirely baleful influence on public debate, is doing his poor man’s version of Freakonomics again over at Slate and attracting an entirely fair amount of opprobrium for doing so (via Matthew). This week, we have the “counterintuitive” “result” (note two different flavours of scare quotes here; the first set are mocking Landsburg for constantly referring to things as “counterintuitive” when they are actually just silly, while the second set is there in order to indicate that his argument is intended to resemble a result from economics, but is no such thing) that turning off the ventilators of patients too poor to pay their medical bills is the right and even the moral thing to do. I don’t think anyone was ever likely to have been convinced by this, but below I present an argument which might help to sort out cases in which economists might have something useful to add to a debate of this kind, from cases like this where they probably don’t.
[click to continue…]
by Daniel on December 16, 2005
Is it me, or is The Economist getting much better? There’s nothing like as much out-and-out arrogance about its pronouncements as there used to be and they seem to be giving opposing views a much fairer treatment. I would say it’s about 40% less pompous than it was at the peak, which is bordering on readability. In the end, though, it is the market which will decide.[1]
Anyway, the rundown they give on the Iraqi elections is excellent and has certainly cleared my thoughts up substantially. (see here for what they looked like when they were unclear). I now think I know enough to make a few predictions about the elections, and if you say that with enough attitude on the word “think” then you’ll realize why The Economist is a dangerous drug that needs to be kept out of the hands of slightly tipsy businessmen on trains.
Update: Well what a bloody washout these predictions turned out to be!
[click to continue…]
by Daniel on December 8, 2005
Scott Burgess at the Daily Ablution blog is in the process of retranslating “The Project” from a French translation published in a Swiss newspaper. Apparently “The Project” is a secret document which outlines the secret plan of the Muslim Brotherhood to infiltrate European institutions, secretly take control of European governments and rule the world. Understandably, Scott is at pains to tell us that “this isn’t a conspiracy theory”, but I think he’s batting on a sticky wicket here; he’s got a theory, and it’s about a conspiracy, so there is no other two-word phrase which describes it more accurately than “conspiracy theory”. Scott himself appears to have a tiny bit of critical distance preserved from this material, but he’s not exactly shying away from the conspiracist interpretation and there are plenty of people in the Daily Ablution comments section who have really gone off at the deep end in the most hilarious fashion possible.
Welcome to the wacky world of conspiracy theories guys is what I say. As a frequent inhabitant of conspiracy mailing lists, can I offer the following advice:
[click to continue…]
by Daniel on December 6, 2005
What with one thing or another, thinking among the thoughtful is now turning to the subject of getting out of Iraq. As someone who opposed getting in there, I just wanted to set down a quick note on an important point; just as I always insisted before the invasion[1] that the question was not “War?” but “this war now?”, it also has to be taken into account that the question now is not “Withdrawal?” but “withdrawal now?”.
[click to continue…]
by Daniel on December 6, 2005
Lots of our American readers complain whenever CT runs cricket coverage. To help “you guys” out, here’s a nice cartoon summary of the rules.
All you have to do is learn French.
by Daniel on November 8, 2005
I am not quite sure why, but I am both too embarrassed with my views on the subject of the riots in France to post them on CT and too pleased with them to resist linking to them elsewhere. Sorry.
by Daniel on October 30, 2005
Sorry this has been a few days in coming, I’ve been tied up. So anyway, is he going to jail or not? My summary advice to both sides would be, don’t get your hopes up. I am getting a clearer picture of what actually went on between 1999 and 2001 with respect to Galloway, Iraq and oil, but there is still a big murky patch of uncertainty. I would also submit that Galloway is correct on one important point; despite the great big smile all over his face on the news, the Presidential hopes of Senator Norm Coleman are probably dead and buried and can’t be redeemed by getting Galloway – the secret is out that he is a flat-track bully who falls apart under pressure and anyone facing him in a debate from here to the end of time will know that. But anyway, what is the new news in the Senate Report [1]?
[1] Actually the majority (ie Republican) staff report from the committee but I do not think this detail is important.
[click to continue…]
by Daniel on October 25, 2005
My putting in the third consecutive CT post on the Bernanke appointment is exacerbated by the fact that it’s only a quick link, but redeemed by the quality of the link. As always, Max Sawicky has his priorities in order:
Best of all is the passing over of the obnoxious Martin Feldstein, who will have to content himself with endless attacks on Social Security from his bunghole at Harvard University
It’s true. Feldstein’s work on Social Security has been, by and large, disgracefully bad (in particular, his claim that SS reduced private saving was based on an error in a computer program; Brad DeLong rather coyly says that these days Feldstein “prefers to stress other points” but he has never really retreated from this claim) and I hope Max continues reminding the world of this fact forever.
[click to continue…]
by Daniel on October 18, 2005
Apparently Sir Ian Blair has no plans to resign. This is good news because, when combined with his statement that he wasn’t going to resign in the immediate aftermath of the Menezes shooting, it raises his “Galbraith Score to 2. It might even be as high as 3 if he said he wasn’t going to resign after he was caught trying to arrange a cover-up of the independent inquiry into the shooting. The trigger score is four, remember (it’s derived from the JKG quote “Anyone who says four times that he won’t resign, will”) so there is plenty of hope that we will be rid of the authoritarian, careerist oaf soon.
by Daniel on October 15, 2005
Tim Lambert has done very good work over the years keeping people honest on the John Lott “More Guns, Less Crime” thesis and on the Lancet study. However, his work on the strange subculture of DDT loons also deserves a bit of publicity.
Basically, there are lots of people out there, mainly the same sort of people who are fans of Stephen Milloy’s “junkscience.com”, who believe that “liberals and environmentalists” are responsible for the deaths of over 50 million people in the third world from malaria because they banned DDT in the 1970s, because they read the book Silent Spring by Rachel Carson. This charge is rubbish from start to finish … as in, DDT wasn’t banned in the 1970s, and using DDT is usually not the best way to prevent malaria. Tim’s DDT archive has the whole damn story.
Why are people so keen on DDT? Don’t know. There’s no compelling economic interest in treating the stuff as if it were a panacea; it’s a commodity chemical which is banned as an agricultural pesticide (in order to avoid creating resistant mosquitoes and compromising its use as an antimalarial) and which has only a niche demand as an antimalarial (because pyrethroid-treated mosquito nets are usually a more cost-effective prevention method). All I can think of is that claiming that environmentalists are responsible for millions of deaths in the third world is a handy way of slagging off environmentalists. One has to say, looking at the calibre of human being pushing the DDT argument in Tim’s archive, for them to cry crocodile tears over the genuine problems of the third world while doing nothing to ameliorate them, simply to fight a domestic political battle, would not exactly be out of character. Nice one Tim for exposing this vile, pernicious rubbish.
by Daniel on September 22, 2005
Hey up everyone, it’s the Prospect Magazine Intellectual of the Year contest!! I’m afraid that everyone at CT forgot to put our forms in (again!) although I see that bloody Yusuf al-Qaradawi did (and Airmiles too). This is a real shame, since I recently became The Most Important Thinker In The World. As Tyler Cowen pointed out in a post on Ray Kurzweil, the previous holder of that title, the secret to being the Most Important Thinker In The World is a mastery of the expected utility rule.
No matter how ludicrous your predictions, if they are sufficiently wildly utopian, then your thinking has a greater expected value than anyone else’s (see here for the general idea). Thus, if Kurzweil reckons that we will upload our consciousness onto software and live for ever as pure energy on the internet, then I say all that and a pony too! Not just any old pony by the way, but a super technonanopony! Which eats racism and shits pure gasoline … on the internet! Oh yeh and we will constantly be having multiple orgasms … and not just the normal kind either (more details to come). You might say that it’s pretty unlikely and I’ve failed to spell out important details, but as long as there is at least some probability that I’m right, then I am more important than Ray Kurzweil to the tune nU^(-rT), where U is the utility of a magic pony, n is the probability I’m right, r is the discount rate and T is the time it will take to sort us all out with one. Keep reading CT folks, because in expected value terms, it is only going to become more important!!
[click to continue…]
by Daniel on September 21, 2005
If something can get worse it will
Is a phrase I’ve often sardonically used
But here’s some grist for Mr Murphy’s mill
I used to be disgusted
Now I’m just amused
Pretty much sums up the way I feel about the Lancet/Iraq debate these days. To be honest all the work has been done; all the big mistakes have been made and corrected, the conspiracy theories have been more or less put to bed and there are really very few people left who believe either a) that things didn’t get worse in Iraq as a result of the invasion or b) that they only got a little bit worse. All that really remains for me to do is to keep on using the term “Kaplan’s Fallacy” to describe the practice of using the mere fact of uncertainty about an estimate to draw conclusions about the direction of the estimation error, in the hope that the bloody dreadful article Fred Kaplan wrote about the Lancet study will stick to him for the rest of his career. But apparently there are still some people out there with a little bit of credibility intact (no really) who think that they can still get away with this one.
Update: Have a look at this from Simon Jenkins in the Guardian. I think it’s quite extraordinary in that a) each of the individual sentences is simply a bald statement of fact and b) the cumulative effect of the whole lot of them is really much more damning than anything George Galloway has come up with.
[click to continue…]
by Daniel on September 17, 2005
Chris asked, quite correctly, where the blogospheric outrage was about the UK government’s current “anti-terrorism” legislation was. I didn’t have any particularly intelligent analysis to add, which is why I haven’t posted so far, but upon reading the bloody thing, I realise that this is hardly an excuse. So here we go.
For Christ’s sakes !! A Labour government (A LABOUR GOVERNMENT!) is trying to pass a law whereby you can sit down at a pub table, spend the evening talking and come away having COMMITTED A CRIMINAL BLOODY OFFENCE!! THIS IS A BLOODY SPEECH CRIME PEOPLE!! THEY ARE QUITE LITERALLY SAYING THAT THEY ARE GOING TO PUT PEOPLE IN JAIL FOR EXPRESSING THEIR POLITICAL VIEWS!!Do I have to start using the f and c words before anyone notices that there is something quite serious going on? I am as concerned as the proprietor of Shot by Both Sides for my long term career path, but this surely has to be more important. THE GOVERNMENT OF THE UNITED KINGDOM OF GREAT BRITAIN AND NORTHERN IRELAND IS PROPOSING TO PUT PEOPLE IN JAIL FOR POLITICAL SPEECH CRIMES!! If anyone is proposing a quick sing-song outside the gates of 10 Downing Street singing “Glory Glory O Bin Laden” I think I am probably up for it. What the by-our-lady hell is going on?!