A hard day’s cargo cult science

by Daniel on July 29, 2007

Good news readers! I’ve gone mad! I don’t know what it was that tipped me over the edge but I’m now a signed up 27%er and I’ve decided to start applying my new grasp of the scientific method! After all, our scientific institutions are being destroyed by the leftist politicised science of global warming and the Lancet study, and that’s just not on. Luckily my cheerful attitude and can-do approach to statistics survived my trip to the dark side so I’ve been hard at work all morning applying the sort of tenacious scientific critique that my new status as a crazy person allows me to carry out with no qualifications whatever.

I started with the UK Census. I’ve always thought that there were maybe a few more, or possibly less, ethnic minorities in Camden than the census said, so I phoned them up and asked for the data. The woman on the end of the line pointed me toward their website and noted that there was quite a lot of county-level data there which might be helpful. I explained that no, I wanted the data, by which I meant the actual census forms. They won’t release the data! Really! I shouted that this was a fundamental building block of the scientific method, and that her sinister refusal to hand over the forms to any random person who asked was the equivalent of the Catholic Church burning Galileo[1]. While she was on the line, I asked for the last month’s death figures for Central London – after all, since she’s the central registering authority for births and deaths, she ought to have them at her fingertips as they must magically update every time a hospital morgue writes a certificate. I think she was in tears by the time she slammed the phone down, so Advantage: Blogosphere!

Next on to the Dow Jones Industrial Average people. Did you know that there are three entire missing days from their figures, which suspiciously enough[2] just “happen” to be September 12-15, 2001???????Q? I suppose we are meant to assume that this “missing cluster” was selected at RANDOM Some chance. Clearly the leftist MSMs of Dow Jones International censored these numbers, because they would have added so much to the variance of the DJIA that we could no longer be sure that it wasn’t 36,000! Perfidy! Wal-Mart are releasing their Q2 earnings numbers next week, or at least I should say “releasing” their “numbers”, because as I found out, when you go down to Bentonville demanding a look at the till rolls, you don’t even get let into the car park. Scientific method, my ASS!

Stay tuned for more science, readers, because until this case of Red Bull runs out, I am going to be a blogoscientific force of nature!

[1] Galileo was not actually burned, but I am now a right wing crazy person, so this kind of factul nitpicking no longer bothers me.

[2] The fonts are a lot more fun on this side of the political divide too.

{ 30 comments }

1

John Emerson 07.29.07 at 1:52 pm

The Dow passed 36,000 some time ago, Daniel. It’s somewhere aroung 42,000 now.

Also, the 27% crazification figure is overpessimistic, since Illinois is infested with liberals. The national figure is above 35%, possibly above 40%, so there’s still hope.

Unfortunately, a major faction of crazies has opted for bipartisanship / nonpartisanship recently. Those people are hard to figure.

2

P O'Neill 07.29.07 at 1:53 pm

You won’t be a made man until you show up at a George W. Bush event to ask this kind of question

Q I personally admire the way you’ve conducted the government and I admire your backbone, where you just stand and take a position. (Applause.) I’m not happy about the influx from Mexico. Seems that far too many came over in waves. I know that during the days of San Jacinto that they were fighting, using rifles and everything, but this is the first time I’ve ever seen an influx like this to try to take over our country. Now then, thirdly, when they do these polls to determine how you’re rated, how come, if they have 1,000 people, they call 750 Democrats and only 25 Republicans? (Laughter and applause.)

3

Tim Worstall 07.29.07 at 2:05 pm

Quite: but then again, there really are dodgy statistics out there which need explaining: like, for example, the claim that 37 million Americans live in poverty. Not quite so, for that is the number below the Federal Poverty Line “before” any of the help they get (in detail, before housing vouchers, the EITC, Medicaid, food stamps etc, after direct cash transfers). European poverty numbers by contrast are almost always quoted after the influence of the tax and benefits systems.
But then my worrying about such things is easily explained: I really am a crazed right winger.

4

Robert 07.29.07 at 2:19 pm

You don’t have to go back to 12-15 Sept 2001 to find oddities in the DJIA. I just looked at the DJIA over the last seven days.

28 July 2007: 0
27 July 2007: 13265.47
26 July 2007: 13473.57
25 July 2007: 13785.79
24 July 2007: 13716.95
23 July 2007: 13851.08
22 July 2007: 0

The mean of the DJIA over the last seven days is 9727.55 with a 95% CI of (-3302,+22757). Dow-Jones is suppressing the fact that there is no valuation to the market!!!!

5

Robert 07.29.07 at 2:38 pm

Omigod. I just remembered: David Kane is a portfolio manager. Coincidence? You be the judge.

6

bi 07.29.07 at 3:14 pm

27%? Isn’t that a Stalinist statistic?

This statistic is wrong. We should totally redo it… until the statistic is free from any Stalinist bias.

(I’m not even entirely making this up. During Google’s interview with Republican candidate Ron Paul, he suggested that we should totally “redo” the statistics on global warming. Goo goo g’joob!)

7

JP Stormcrow 07.29.07 at 3:37 pm

[1] Galileo was not actually burned, but I am now a right wing crazy person, so this kind of factul[sic] nitpicking no longer bothers me.

Keep pushing a little further brother and soon you will have access to truths such as:

…the satanic Sarpi’s lackey Galileo had attempted sodomic rape on the body of Kepler’s discoveries of the Creator’s universal physical principle.

8

Fr. 07.29.07 at 3:40 pm

… or use Giordano Bruno in replacement?

9

David Kane 07.29.07 at 4:33 pm

LOL, as the youngsters say.

But as to data release, could you explain why Roberts et al release their data to most who ask for it, except for Michael Spagat and his co-authors?

10

J— 07.29.07 at 4:38 pm

But the Church did burn Galileo. I read about it at Ultramontane Media Mythbusters.

11

MattF 07.29.07 at 4:49 pm

You’re a paragon of sanity. Watch the banana video.

12

Aaron Swartz 07.29.07 at 4:56 pm

Thank you so much, Mr. Davies. I can’t tell you how wonderful it is to have you on the team.

Do you think that next week you can explain to us how the One Honest Science — neoclassical economics — tells us that any rational person will stay in Iraq indefinitely no matter how bad the news gets? Extra thanks if you can throw in some of those number pointy things that make it look all serious and extrascientific.

13

John Emerson 07.29.07 at 5:02 pm

I would guess it’s probably because they think Spagat is a fraud or an incompetent.

14

David Brooks 07.29.07 at 5:22 pm

I don’t really understand your post, but I like its tone. Can I publish it under my own name? I’d really like to stick it to that smartypants ‘economist’ Krugman.

15

Barry 07.29.07 at 6:35 pm

John Emerson, you have commited left-wing rationalistic deviationalism; please report to the Malkinization chamber.

Kane, so far there has been no proof[1] whatsoever that Spagot has asked for the data, or any proof[2] that Roberts has not fulfilled this request that we don’t admit ever happened.

Therefore, we have no choice but to brand you a liar.

-Barry
[1] from any source which we deign to trust.
[2] ” ” ” ” “

16

John Quiggin 07.30.07 at 3:54 am

“Not quite so, for that is the number below the Federal Poverty Line “before” any of the help they get (in detail, before housing vouchers, the EITC, Medicaid, food stamps etc, after direct cash transfers). European poverty numbers by contrast are almost always quoted after the influence of the tax and benefits systems.”

AFAIK, standard European numbers don’t include benefits of public health care or public housing, even though health care at least is much better than what’s available for the poor in the US. I’m also surprised by the claim that US numbers exclude EITC. Assuming the numbers are net of tax and benefits, how does this work?

17

shub-negrorath 07.30.07 at 4:55 am

#11: But unlike Daniel, the banana guy isn’t kidding.

18

ejh 07.30.07 at 7:52 am

Mr Worstall seems to have accidentally overloooked that what statistics say depends opn the reason why they are collated and the reason those particularly statistics are collated is to determine which people are eligible for the benefits he describes. Other statistics which differ in kind may be collated for somewhat different purposes.

He might respond to the effect that yes, this may be so, but doesn’t that mean that having received these benefits the people in receipt are no longer to be considered poor? To which the answer is yes, of course, in the sense that people whose poverty entitled them to a place in the workhouse were, on arrival in the workhouse, no longer poor.

19

bad Jim 07.30.07 at 9:58 am

And to think that our troops in Iraq have been teaching abstinence (or is it gun control?) for four years now, and the citizenry still hasn’t caught on. What do we have to do to penetrate their thick skulls? Apart from shooting them, of course, which we’ve already tried.

20

Tim Worstall 07.30.07 at 11:10 am

“I’m also surprised by the claim that US numbers exclude EITC. Assuming the numbers are net of tax and benefits, how does this work?”

John, it’s there on the Census site. US Federal Poverty Line is calculated before tax (thus, the EITC, which works through the tax system is not included) and before benefits in kind, but after direct cash transfers (ie, traditional welfare).

“standard European numbers don’t include benefits of public health care or public housing”

UK numbers do include housing benefit, but not social housing. So, no, they’re not perfect either.

“Mr Worstall seems to have accidentally overloooked that what statistics say depends opn the reason why they are collated and the reason those particularly statistics are collated is to determine which people are eligible for the benefits he describes. Other statistics which differ in kind may be collated for somewhat different purposes.”

I agree. But the thing is, the US does not collect the statistics (on a regular basis, the last set were from 2004, when they looked at alternative measures of poverty) on what is the incidence of poverty after whatever redistribution there has been.

My complaint is not that the Fed Pov Line is drawn so as to identify those who need help. But that it is used politically to argue for more help: it may well be that the poor need more help (I certainly think so) but the correct numbers used to measure this would be those after the influence of the current poverty amelioration efforts.

At the risk of boring people with statistical details: this is, at least to my mind, important re the John Edwards poverty plans. He argues for more housing vouchers, a higher EITC amongst other things. He claims this will reduce poverty. I think he’s probably right. However, given the way that poverty is measured in the US (and JE uses this definition himself, the number below the FPL) it will have no effect whatsoever on the number recorded as being in poverty. Because neither the EITC nor housing vouchers are included in that FPL calculation.

We can go further: as many have noted, there’s not been much reduction in poverty in the US in recent decades. As others have also noted, Federal spending on poverty reduction has risen strongly in recent decades. We thus have from the one side, those stating that see, spending money on reducing poverty doesn’t work. On the other, we have people pointing to extant poverty and claiming that much more needs to be done.

However, since the early 1970s there’s been a (bipartisan) move from direct cash payments of welfare to payments in kind and the EITC as poverty alleviation methods. So spending on poverty alleviation has gone up: but the headline poverty number hasn’t moved because we don’t include that spending in our calculations of the poverty numbers.

Quoting oneself is of course horribly egotistical but from last year:
http://www.tcsdaily.com/article.aspx?id=072506B

“My apologies for working with old numbers but that cash assistance in 1979 was $35 billion, in 1986, $32 billion. The non-cash part was $97 billion in 1979 and $136 billion in 1986 (figures in 1986 dollars). The craziness is that the 32 and 35 numbers are included in the incomes before we calculate those who are in poverty. The 97 and 136 are not. So we’ve actually had a rise of $36 billion in spending upon the poor but (that portion of their incomes that come from the Feds anyway) the incomes of the poor have fallen by $3 billion. Given the way these numbers are calculated, the way the official figures work, no wonder we haven’t been able to eradicate poverty.”

I’m not arguing (here, anyway) for either more or less support for the poor. Only that any decision either way should be taken after we look at the effects of current spending: which current statistics do not do.

21

Barry 07.30.07 at 11:23 am

Well, quoting from TCS certainly fits in with cargo cult science.

22

Tim Worstall 07.30.07 at 1:18 pm

barry # 21: I think I already made that joke up above. I’m also quoting myself which obvisouly makes it even worse.

23

JP Stormcrow 07.30.07 at 2:22 pm

I do think that as far as cults (or religions in general) go, you can argue that “Cargo Cults” were among the most successful in history in actually delivering on their goals. It is not unreasonable to think that the practices in question led to an increase in cargo for their practitioners via the agency of increased interest and visits from people from “cargo rich” parts of the world.

I have no data to support this argument, but I’m right.

24

Barry 07.30.07 at 3:07 pm

jp, you’re righter than right, because your rightness is pure, uncontaminated by data.

25

ajay 07.30.07 at 4:37 pm

jp stormcrow: actually, the most successful religion in the world in the sense of delivering on its goals was the Aztec cult of the sun. The Aztecs sacrificed thousands of people in order to make the sun come up the next morning. Sure enough…

26

ejh 07.30.07 at 4:58 pm

The Observer once ran a headline saying something like “Corruption still present in business, says survey”. I wrote them a letter asking if they proposed to do a follow-up story saying “sun rises in morning” or “Earth spins on axis”. I thought this was pretty clever until my landlord pointed out that those were, in fact, the same story.

27

John Quiggin 07.31.07 at 2:29 am

Tim, if the poverty measure excludes tax but includes cash benefits, then this is going to understate poverty relative to the European approach, except for people for whom the EITC exceeds all tax payments (can’t tell, but there don’t seem to be many of these).

Of course, I agree that it’s a bad measure and that it distorts discussion of EITC in particualr.

28

Tim Worstall 07.31.07 at 7:33 am

John, not quite sure of that: US income tax starts at a pretty high level. Most of those in poverty would be paying FICA (if at all) but not income tax ($7,500 personal allowance, tax is 10 or 15% above that, poverty level for an individual is some $10,000 or so.)
I had this discussion with Max Sawicky a few weeks back: are there actually any comparisons of European v. US poverty rates? It would appear that there aren’t. For over and above the problems detailed above (and your final sentence is the very reason that I noted it in the first place. When discussing the EITC and poverty, it would be useful to actually measure the effect of the EITC on poverty.) we’ve got two entirely different systems of measurement.
The US system is broadly a measure of absolute poverty (a particular lifestyle, in the 60s, since uprated for consumer inflation, but not for real wages) as against the usual European method (also the OECD recommended) of relative poverty (60% of median household income, adjusted for household size and housing costs isn’t it?).
I’m pretty sure you’re right, that poverty in the US would be higher using the European measurement systems (a higher Gini makes that almost inevitable) but I’ve not been able to find any papers that actually try to calculate it. If you know of any please do let me know, I’d love to see them.

29

dan 07.31.07 at 8:06 pm

David Brady provides an extensive review of the poverty literature and then uses the Luxembourg Income Study (LIS) to produce comparable estimates using a relative definition of poverty. His work here is particularly interesting because he also makes use of Amartya Sens innovations in poverty measurement.

The cite is 2003. “Rethinking the Sociological Measurement of Poverty”. Social Forces 81:3. I know I’ve seen some others as well, but can’t seem to find them at the moment.

A couple of asides–it would be very hard to convince anyone in the US of the utility of a 60% of median measure of poverty. The current US measure corresponds pretty closely to a 40% of median measure (assuming proper household adjustments). Which means the new measure classifies a much larger number of people as in poverty than the old. Also, while it’s certainly not a common procedure, it’s possible to account for the effects of the EITC and taxation. The workaround isn’t perfect (calculating the EITC is easy, income tax not so much–and it’s also pretty inaccurate) but it can certainly be done.

30

mjfgates 07.31.07 at 9:11 pm

By the time you’re getting the EITC, you’re paying zero income tax; whether your total taxes will be negative or not varies from state to state. When I added it all up for myself in 2006, EITC + deductions just about paid for every tax-like thing I have to pay, sales taxes, Social Security, workman’s comp, etc.

I got a better job this year, though, so this time round I’ll actually have to PAY taxes. Add it all up and it only comes to about ten grand more in my pocket than when I was underemployed, snif snif.

Comments on this entry are closed.