The elephant's trunk

Via David Appell, I came across this marvellous quote from Freeman Dyson

In desperation I asked Fermi whether he was not impressed by the agreement between our calculated numbers and his measured numbers. He replied, “How many arbitrary parameters did you use for your calculations?” I thought for a moment about our cut-off procedures and said, “Four.” He said, “I remember my friend Johnny von Neumann used to say, with four parameters I can fit an elephant, and with five I can make him wiggle his trunk.”

It came to mind when I read this story in the NYT with the introductory claim What really stimulates economic growth is whether you believe in an afterlife — especially hell.The report is of some estimations done by Rachel M. McCleary and Robert J. Barro (the story notes that the two are married) published in American Sociological Review.

Barro is probably the biggest name in the field of cross-country growth regressions (a field in which I’ve also dabbled), and I’m sure he’s aware that thousands of these regressions have been run and that, with very limited exceptions, results that particular factors are conducive to growth have proved highly fragile. I haven’t read the paper, so for all I know, the results have been checked for robustness in every possible way. But my eyebrows went up when I saw this para

Oddly enough, the research also showed that at a certain point, increases in church, mosque and synagogue attendance tended to depress economic growth. Mr. Barro, a renowned economist, and Ms. McCleary, a lecturer in Harvard’s government department, theorized that larger attendance figures could mean that religious institutions were using up a disproportionate share of resources.

What this means is that at least two parameters have been used in fitting growth to religiosity and that the two have opposite signs – most likely it’s some sort of quadratic. In my experience, there’s always at least one arbitrary choice made in the pretesting of these models (for example once you have a quadratic, the scaling of variables becomes critical). That gives three free parameters, if not more.

I’m not John von Neumann, but with two parameters I can fit a dromedary and with three I can do a Bactrian camel.

New discoveries in evolutionary psychology

I just got the latest issue of Scientific American, and noted with interest the Table of Contents, in which the Skeptic column promised an evolutionary explanation of the mutiny on the Bounty. I vaguely expected the usual stuff about alpha and beta males or somesuch, but I found that the ev psych boffins have come up with a startling new discovery. Young men like having sex. At this point the mathematics and biochemistry get a bit complicated for me (oxytocin is in there somewhere), but apparently this has something to do with the survival of the species.

Even more startling, though, is the fact that

Although Bligh preceded Charles Darwin by nearly a century,

he managed to anticipate this discovery. Who would have thought that a former governor of New South Wales (and not a successful one) would share with EO Wilson and Stephen Pinker the honour of founding evolutionary psychology? In Bligh’s words

I can only conjecture that they have Idealy assured themselves of a more happy life among the Otaheitians than they could possibly have in England, which joined to some Female connections has most likely been the leading cause of the whole business.

Delivery times are somewhat strange here in the Antipodes, and I thought perhaps I had an advance copy of the April edition, but the cover says February.

Second-best blog in Queensland: Collect $10

There seem to be “Best blog” awards going on all over the place, but the only one where this blog a was serious contender was the Australian Blog Awards over at Keks. The contest was run on proper Australian lines with optional preferential voting, and when the preferences were all distributed, 60 per cent of the “two-blog preferred vote” for best Queensland blog went to 85 George Street, with this blog as runner-up.

Although not everyone is happy about awards and so on, I think it’s all good fun (It would have been even more fun if I’d won something!). Thanks to vlado at Keks for taking the trouble to run this, and thanks to everyone who voted for me or who took the trouble to vote at all.

Big congratulations also to Gianna, who got best NSW blog, Meika (Tasmania), Troppo Armadillo (NT), Ubersportingpundit, one vote ahead of Gary Sauer-Thompson in SA, and Sam Ward and Robert Corr in WA. I realise that I’ve missed out the two great Tims of Australian blogging and quite a few others, but if you go back to Keks you can see the entire list. As with the Laughing Clowns, just about every player has won a prize, which is as it should be.

Wendy James posts the thoughts

Wendy James posts the thoughts on public schools of pseudonymous teacher S. Whiplash, who says

If the figures are to be believed, increasing numbers of Australian parents are choosing to send their children to private schools. I believe this is happening totally – well, maybe not totally, but damn near – because parents are unhappy with the values being taught in public schools. Some of these values are taught overtly and some covertly. Either way, parents don’t like the values package on offer and are voting with their children’s feet.

As it happens, my opinion piece in yesterdays Financial Review (subscription required)was on this very topic, making the point that economists would look at the issue rather differently

Consider first the premise, shared by Howard and many of his critics, that the shift in enrolments from public to private schools must reflect increasing dissatisfaction with the public system. An economic appraisal suggests a much less abstract explanation.

Thanks to changes in Commonwealth government policy, subsidies to private education have been steadily increasing. Meanwhile the effective subsidy to publicly educated students has remained constant or declined in recent years. Standard economic analysis suggests that when a service is subsidised, its consumption will increase.

The analysis even works when comparing Catholic and other independent schools. Under Labor, the Catholic system received fairly generous assistance, but aid to the wealthier independent schools was limited. The Howard government has greatly increased aid to the wealthiest schools and enrolments have followed, with both government and Catholic schools losing market share in recent years.

From this perspective, in fact, the surprise is that the increase in attendance at private schools has been so small. In 1963, before the Menzies government began the provision of government aid to private schools, around 24 per cent of students attended non-government schools. After 40 years of steadily increasing public assistance, the non-government share has reached only 32 per cent. This suggests either that parental preference for non-government schools is very weak, or that the perceived advantages of private education have been declining over time.

I should observe that one reason for high attendence at private schools before 1963 was the effective subsidy provided by the voluntary labour of members of religious orders in Catholic schools. The gradual disappearance of this group and its replacement by lay teachers, paid out of public subsidies, has greatly reduced the differences between Catholic and government schools.

Manners and political correctness

In my experience there is a close to 100 per cent correlation between the stated belief that society is suffering from a decline in “civility” and a willingness to proclaim that we are all being oppressed by “political correctness”. John Howard neatly illustrates this. A week or two ago, he was denouncing public schools as hotbeds of political correctness, and the excessive concern with offending religious minorities that (allegedly) led to the curtailment of Christmas celebrations. Now he’s calling for more civility.

The common analysis underlying both demands for “political correctness” (this actual phrase was never used, except jocularly as far as I know, until critics seized on it, but terms such as “sensitivity” or “inclusive language” cover much the same ground) and for “civility”, is that offensive words give rise to offensive acts. In both cases, there’s some ambiguity over whether the problem is with the offence to the recipient or with the reinforcement of the hostile/prejudiced attitudes of the speaker, but the central claim is that modes of speech are an appropriate subject of concern and that some form of government action to encourage more socially appropriate modes of speech, ranging from subtle pressure to direct coercion, is desirable. The only difference between the two positions is that they have different lists of inappropriate words.

I don’t have a sharply defined position on any of this, except that I find people who think that being “politically incorrect” is exceptionally brave and witty to be among the most tiresome of bores. I doubt that changes in speech will, of themselves, produce changes in attitudes. The obvious evidence for this is the rate at which euphemisms wear out and become as offensive as the terms they replaced (for example, ‘handicapped’ for ‘crippled’). On the other hand, I think there’s a lot to be said for avoiding offensive words and forms of speech and can see a place for (tightly drafted and cautiously applied) laws prohibiting or penalising various forms of collective defamation.

[Posted with ecto]

Belated congratulations

To Steve Waugh on being selected as Australian of the Year. Commentary on the award focused on the fact that Waugh was the third Australian cricket captain in a row to get it. The implied critique is that, if the award is thought of as rewarding excellence in some particular sphere, sport in general, and cricket in particular, get much more than their fair share.

My view on this is that, assuming the award is some sort of all-round commendation of a prominent Australian, Waugh is considerably more deserving than his two predecessors. His charitable work in India, and all-round good behavior seems to go well beyond the perfunctory good citizenship expected of sporting role models. I hope that Waugh’s award will set a high benchmark and that the committee (or whoever) will think long and hard before making another award based on nothing more substantial than a good batting average.

Top-up fees (crossposted at Crooked Timber)

I’m not clear enough on the workings of the British Parliament to know whether Blair’s 5-vote win on the second reading of his education bill means that the political fight is over, but I thought I’d have my say anyway. The core elements of the bill are a substantial increase in fees, the right of universities to charge variable “top-up fees” and the introduction of a HECS-style repayment mechanism using the tax system. Thus it’s like a combination of all the education financing changes in Australia from the late 1980s, when HECS came in, to the recent Nelson package. Not surprisingly, I like some parts of it, and dislike others.

First, I’ll respond to other CT bloggers who’ve discussed this issue. Chris primarily makes the argument that, given that money isn’t going to come from anywhere else, or on any other terms, it’s better to take what’s on offer than to refuse on the basis that the terms are bad ones. I suppose I agree with this, but it’s not a helpful basis on which to discuss policy. Assuming you don’t want the Tories back, the same argument could be used for acquiescence in whatever policy Blair chooses to propose. Chris also dismisses concerns about variable fees, and I’ll return to this.

Daniel argues on risk grounds against the repayment mechanism (borrowed from the Australian HECS scheme) and, in my view, gets the risk analysis wrong. For precisely the reasons he outlines for not using NPV rules in assessing the effects of fees, the insurance implicit in the provision that no repayment is required until/unless earnings exceed some percentage of average earnings is considerably more valuable than he suggests. Assuming the proportion is set to give a level higher than the average earnings of non-graduates, it makes education a one-way bet. If you win, by earning more than you would have expected otherwise, you pay back some of your winnings. If you lose, you pay nothing. I don’t know what the actual proportion is, so I should stress that my support for the repayment scheme depends critically on this variable – in the absence of a high threshold substantial insurance, Daniel’s analysis is correct.

The critical sticking point, though, is not the level of fees but the principle of variable fees. If this provision had been dropped, it seems clear that the rest of the package would have passed fairly easily. The claim that these are not the same variable fees that were specifically excluded in the manifesto is nonsense, and the determination with which Blair and Clarke have stuck to them shows this.
Read More »