Blue Thunder (A Modern Bluey Brink)

There once was a trendy who scorned worldly wealth
And devoted his life to a search for good health,
He spent all his money at his local health store
Took potions in gallons and pills by the score

But sad to relate every time that it rained
His house it was flooded from roots in the drains
To cure it the landlord gave him a pill
To flush down his dunny, them tree roots to kill

Next morning he followed his usual routine
By swallowing tablets in number thirteen
But somehow or other one extra crept in
It was big, it was blue, it was bitter as sin

This pill was so big that it stuck in his throat
It made him to splutter and it made him to choke
But he swallowed it down like he knew that he should
If it tasted that vile, well it had to be good

Now soon he was stricken with aches and with pains
As the pill started working to clear out his drains
He was stuck on the dunny for a night and a day
Weighed fifteen pounds less when he staggered away

Well it made him feel weak but it made him feel pure
So he told all his friends of his miracle cure
They asked him the secret but he wouldn’t tell
After buying for years he had something to sell

And now he is running his own health food shop
Selling soya bean sausage and Gaylord’s green glop
And his own extra-specialty, cure for all ills
They’re bio-dynamical Blue Thunder Pills

I just thought I’d record this statement (made in connection with the discovery of empty chemical warheads in Iraq) by Andrew Sullivan for future reference

There can be no further excuses. Saddam had one absolutely last chance and he lied. If we do not go to war now, then Bush, in turn, will have been shown to have lied in his countless statements declaring zero tolerance for future violations.

Given Sullivan’s generally hagiographic style in relation to Bush, even a future conditional use of the word ‘lie’ is striking.

While I’m on this point, today’s Age reproduces a piece from the LA Times giving some factual background, including the point:

Dozens of rockets are usually fired at once, to try to ensure that enemy troops are enveloped in a thick concentration of chemicals.

This seems relevant in assessing the status of the 12 warheads found so far.

Following my posts the Windschuttle controversy, I promised to put forward some ideas on the current policy problems facing Aboriginal Australians, and particularly the problem of economic development. It’s always problematic for white ‘experts’ to tell black communities what to do and I want to make it clear that I’m not trying to do this. Although I have given economic advice to Aboriginal organisations on a range of issues, I don’t regard myself as an expert on the problems facing Aboriginal communities. My perspective on the issue comes more from a consideration of the general economic problems of rural Australia and particularly the general decline in population and employment.

Discussion of economic policy for rural and remote Aboriginal communities often poses a dichotomy between passive reliance on government welfare and economic independence. The assumption is that reliance on government is inherently demoralising.

I have come to the conclusion that, at least in the terms in which economic independence is commonly understood, this kind of thinking is mistaken. Against a background of generally declining demand for labour in rural and remote Australia, it is unlikely that the majority of Aboriginal communities will ever be economically independent in the sense that they produce enough in terms of market goods and services, to ‘pay their way’ in a free-market economy, even with an inflow of capital income from land and mineral rights. Aboriginal communities have all the economic problems facing declining country towns as well as the additional difficulties arising from a century or more of dispossession and discrimination. If most country towns can’t find a way to hold enough jobs for a generally shrinking population, it seems impossible that Aboriginal communities, many with growing populations, can ever ‘pay their way’.

Having said that, I will observe that for most of the 20th century, large sections of the Australian workforce did not ‘pay their way’ either, relying instead on tariff protection or subsidy schemes. This did not seem to do much harm to the self-respect of Australian workers. It’s my view that if ‘practical reconciliation’ is taken to include the objective of generating sustainable employment in rural and remote Aboriginal communities, we have to accept that employment will be sustained only with permanent government support, just as was true of large sections of manufacturing industry. The obvious approach is permanent wage subsidies, though in some cases output subsidies to enterprises based in Aboriginal communities might also work.

The closest approach so far has been the Community Employment Development Program under which communities can use social security benefits to pay for useful projects. This scheme (which has elements of ‘work for the dole’) has been possible only because it has been treated as welfare rather than as a labour market program. The idea of a permanent wage subsidy, explicitly confined to rural and remote Aboriginal communities breaks so many policy taboos that it is unlikely to make it on to the policy agenda for a long time. But I can’t see any alternative approach that will generate substantial numbers of permanent jobs.

Can any scientifically literate readers comment on thisABC Report, which says

research reported in the science journal Nature challenges basic beliefs of evolution – that wings evolved only once in insects and that if a trait is lost is cannot be regained.

It also opens a new direction for research because it shows that once a complex figure has evolved it can be maintained over long evolutionary period even if it is not apparent on the outside.

Over a 50 million year period, even though the stick insects did not have wings the genes for creating them appeared to have been maintained.

“The remarkable thing was that they had the ability to generate wings when they needed them,” Mr Whiting explained.

I was under the impression that the standard mechanism in the loss of things like wings was the selection of genes that ‘turned off’ crucial steps in development. If so, it would obviously be easier to reverse this process than to evolve wings for the first time. Am I wrong, or is the article over-dramatizing things?

UpdateThe BG strikes yet again. Brad DeLong has an almost identical reaction with a reference to “Darwin’s Radio ” – must read this.

I was going to comment on the disastrous decision of the US Supreme Court in Ashcroft vs Eldred, upholding new laws which amount to Copyright Perpetuity. I thought I’d point out that Australia still has a more reasonable position where copyright last 50 years after the author’s death, speculate on the implications for the proposed Free Trade Agreement and wrap up with a link to Kim Weatherall and a pithy comment. When I looked at her site, I found

Another thing we can look to, I think, in the current round of Free Trade Negotiations between Australia and the United States, is pressure from the States, freed from the spectre of constitutional challenge, to Australia to extend its copyright terms to match those of the United States. So perhaps what we REALLY need is some pressure back here in Australia.

So my pithy comment is “Me too!”

Ozplogger James Morrow has a piece in Wednesday’s Oz, saying

If a gambler sitting at the Crown Casino in Melbourne knew that the Government would cover his losses, no matter what, he’d have no incentive not to draw to an inside straight or hit on 17. Likewise, a board of directors, if they believe their business is politically “too big to fail” and could qualify for a government bailout, will surely be more inclined to take the sort of big risks that would in the long run cost a gambler his house and car and livelihood.

And though a casino is able to force a loser to pay his debts, in the corporate cowboy world of Adler and Williams, losses go unpaid while the gamblers walk around with huge personal fortunes.

I agree with the sentiment and with Morrow’s concerns about ‘moral hazard’, but I think he has aimed at the wrong target.

It’s not government bailouts that let Adler, Williams, Rich and others walk away from failed companies with their personal fortunes intact. Under the institution of limited liability, this is the rule, not the exception. Assuming the company is not preserved as a going concern, government bailouts of the kind we saw with HIH don’t give any additional help to the managers and shareholders, only the creditors (in this case, policyholders, and in other cases employees).

James writes for Reason and there are quite a few bloggers of a libertarian bent. I’d be interested to read their views on limited liability and, for that matter, personal bankruptcy.

Update As I hoped, this post has generated a lively comments thread. Come and have your say – I’m enjoying the bloggers’ privilege of asking a rhetorical question without revealing my own answer for a while.

PS I forgot to thank Gareth Parker, whose link I followed.

I’ve been arguing for some years that (re)nationalisation should be put back on the policy agenda. The starting point is in reversing obviously failed privatisations, such as that of British Rail. Following the replacement of the private Railtrack company (which owned the network) by the quasi-public Network Rail The Guardian reports the partial renationalisation of track maintenance. More precisely, Network Rail has sacked the private contractors who were doing this job on one of the most accident-prone stretches of track and taken it back in-house.

Network Rail’s announcement was the latest in a series of moves to increase central control over the railways, prompting suggestions of “creeping renationalisation”. In November the strategic rail authority announced that train operators would be moved on to shorter, stricter franchises.

The privatisation of maintenance was at the centre of Ken Loach’s great film The Navigators, which I reviewed for the Canberra Times (my only venture into film reviewing so far). In the review, which is available at Australian Policy Online, I observed

The tendency to shave safety margins an ever-present feature of the privatisation process. It reflects a shift of power from engineers to accountants, and more generally from a production objective to a profit objective.

APO has lots more good stuff and a very nice-looking site.

Ever since Bush’s turn to the UN, I’ve been debating the significance of UN resolution 1441 on Iraq with other bloggers, notably including Steven Den Beste, who initially thought Bush was knuckling under, then changed his mind. When the first draft was announced, that it was, in effect, a one-and-a-half resolution I observed that it effectively required a second UNSC vote

All that’s left in the reported draft resolution, and in the statements of the UK and US governments, is that, if inspectors report obstruction to the UNSC, the US and UK will not necessarily accept a veto on military action cast by, say, France.

Although, the US Administration is still resisting this interpretation, Tony Blair has now accepted it almost word for word

Mr. Blair said, “Of course we all want a second U.N. resolution. I believe we will get one.

But he added, “Where there is an unreasonable veto put down, we will not rule out action.

On this issue, Blair’s interpretation is more important than Bush’s. The US Administration has already made up its mind that it wants a war, whereas the British government has not.

Of course, all of this may turn out to be academic (in the pejorative sense) if the latest reported discovery of shells for chemical weapons turns out to provide the much-sought-after ‘smoking gun’. On the other hand, the precise interpretation of 1441 will be very important if the inspectors turn up incriminating, but not conclusive, evidence against Saddam. It’s still too early to tell on this one.

My Op-Ed piece in today’s Financial Review (subscription required) is on unemployment. A short grab:

At no time since the election of the current government has unemployment been an issue of real concern. Second-order trivia like the GST and waterfront reform have had far more attention. And, sadly, the Australian public has become inured to chronic mass unemployment. In the absence of a severe economic downturn, the government will pay no real political price for its worst policy failure.

Also, today the Committee for the Economic Development of Australia released a review of Australian experience with privatisation. You can download the <a Introduction here (PDF). Most of the contributors are broadly sympathetic to privatisation, though less so in the case of regulated monopolies. I argue, as I have done for some time, that we should be looking at renationalisation in many cases.

Most interesting to me was a study by Jonathan Kelly and Joanna Sikora showing that public opinion against privatisation has hardened steadily over time. In 1986, views on the privatisation of Telstra were about evenly divided. By 2002, 70 per cent were opposed and only 16 per cent in favour. Similar views apply even to firms like the Commonwealth Bank and Qantas that have been privatised for years, and opposition is even stronger in the case of Australia Post, the only business in the study still in full public ownership.

It’s difficult to attribute this to emotional attachments to Aussie icons, in view of the fact that people were quite willing to contemplate privatisation in the 1980s. Advocates of privatisation hoped that experience would lead people to accept it. In fact, the reverse has been the case, here as in Britain and New Zealand. People have experienced privatisation and they don’t like it.

In view of the rhetoric about elites we’ve heard so much about lately, it’s worth pointing out that the divide between elite and popular opinion is far sharper here than in the case of asylum-seekers.

‘Data mining’ is an interesting term. It’s used very positively in some academic circles, such as departments of marketing, and very negatively in others, most notably departments of economics. The term refers to the use of clever automatedsearch techniques to discover putatively significant relationships in large data sets.

The paradigm example, though a very old-fashioned one is ‘stepwise regression’. You take a variable of interest then set up a multivariate regression. The computer then tries out all the other variables in the data set one at a time. If the variable comes up significant, it stays in, otherwise it’s dropped. In the end you have what is, arguably, the best possible regression.

Economists were early and enthusiastic users of stepwise regression, but they rapidly became disillusioned. To see the problem, consider the simpler case of testing correlations. Suppose, in a given dataset you find that consumption of restaurant meals is positively correlated with education. This correlation might have arisen by chance or it might reflect a real causal relationship of some kind (not necessarily a direct or obvious one). The standard statistical test involves determining how likely it is that you would have seen the observed correlation if there was in fact no relationship. If this probability is lower than, say, 5 per cent, you say that the relationship is statistically significant.

Now suppose you have a data set with 10 variables. That makes 45 (=10*9/2) distinct pairs you can test. Just by chance you’d expect two or three correlations that appear statistically significant correlations. So if your only goal is to find a significant relationship that you can turn into a publication, this strategy works wonders.

But perhaps you have views about the ‘right’ sign of the correlation, perhaps based on some economic theory or political viewpoint. On average, half of all random correlations will have the ‘wrong’ sign, but you can at expect to find at least one ‘right-signed’ and statistically significant correlation in a set of 10 variables. So, if data mining is extensive enough, the usual statistical checks on spurious results become worthless.

In principle, there is a simple solution to this problem, reflecting Popper’s distinction between the context of discovery and the context of justification. There’s nothing wrong with using data mining as a method of discovery, to suggest testable hypotheses. Once you have a testable hypothesis, you can discard the data set you started with and test the hypothesis on new data untainted by the process of ‘pretesting’ that you applied to the original data set.

Unfortunately, at least for economists, it’s not that simple. Data is scarce and expensive. Moreover, no-one gets their specification right first time, as the simple testing model would require. Inevitably, therefore, there has to be some exploration (mining) of the data before hypotheses are tested. As a result, statistical tests of significance never mean precisely what they are supposed to.

In practice, there’s not much that can be done except to rely on the honesty of investigators in reporting the procedures they went through before settling on the model they estimate. If the results are interesting enough, someone will find another data set to check or will wait for new data to allow ‘out of sample’ testing. Some models survive this stringent testing, but many do not.

I don’t know how the marketing guys solve this problem. Perhaps their budgets are so large that they can discard used data sets like disposable syringes, never infecting their analysis with the virus of pretesting. Or perhaps they don’t know or don’t care.

Update Kevin Drum at CalPundit gives the perspective of a marketing guy, with lots of interesting points (for example, loyalty programs are there to collect data for mining). He doesn’t accept my main point and raises the dreaded “B” word – Bayesian.

For those familiar with debates among statisticians, this is the point at which things typically become both heated and incomprehensible (just like a lot of blogs, really). A real challenge, which I may tackle at some point, is to explain the Bayesian concept of statistical reasoning in ordinary language. For the moment, though, I’ll just agree that a debate like the one over data mining ultimately makes sense only if it’s cast in Bayesian terms, that is, with a discussion of the beliefs we hold before we begin the statistical analysis.