Quote of the day: On knowledge...

From Don Boudreaux at Cafe Hayek:

Too many are the people who, having mastered econometrics and gathered lots of data, wrongly suppose themselves thereby to possess knowledge. Likewise, too many are the people who, having mastered mathematics and memorized the mechanics of lots of theoretical models, wrongly suppose themselves thereby to possess knowledge. Too rare are the people who correctly understand that, no matter how smart they are and how much they might genuinely learn about economics, econometrics, and ‘the data, neither they nor others can ever hope to come close to knowing the details of economic reality in the same way that, say, a physicist can know the details of some physical material under his or her investigation.

What we call facts (and they aren't always facts, especially in social sciences) do not constitute knowledge. That knowledge comes from thinking about those facts and what they mean, what they tell us about the world.

....

Questioning the location of Macclesfield: the curse of fact checking

****

It was one of those Sunday lunchtime family discussions, the ones that gradually descend into pointless row. And for reasons lost in the mist of time the question of Macclesfield's location arose.

"Macclesfield's in Cheshire."

"No, Macclesfield's in Greater Manchester."

For a minute or two the discussion continues in this vein. I stomp from the room pick up the atlas, find the relevant page and point to the map.

"There you are! Macclesfield is definitely in Cheshire."

A momentary hiatus in the discussion.

"No, the boundaries have changed since that map was published. Macclesfield is now in Greater Manchester."

The discussion continued, now escalated into a row about people who don't believe things even when they're up to the elbow in a spear wound.

Today, in our political discourse we have become obsessed with facts. Or rather with a thing called "fact checking". Rather than engage with reasoned argument we resort to Google and crawl all over the offending comments seeking an error.

At one level this is yah boo - "you've got a date wrong so everything else you say has to be wrong". Or worse still minor grammar errors or a misspelling are fingered - "you can't even write English, why should I believe anything you say."

At another level the correcting of facts is more relevant  - what we might call the 'Location of Macclesfield' question. This might be an understandable confusion between 'income' and 'earnings', a mix-up over two numbers (say illegal immigrants and asylum seekers) or using out-of-date information such as an earlier rate for minimum wage or some or other benefit.

Now it may be the case that one of other of these errors invalidates the argument being made (Macclesfield is in Greater Manchester) but the simple identification of the error isn't enough. You have to demonstrate that, with this error corrected, the argument no longer stands (Macclesfield is in Cheshire). If the argument still stands then the corrected error is just that, a corrected error and not a deal breaker.

"Proving" that someone is wrong by pointing out errors in their facts is great fun if you're the sort of person who doesn't mind getting into a stupid row about the location of Macclesfield ("guilty as charged m'lud") but, as the basis for political discourse it's only a marginal improvement on "I'm right and you're wrong". Sadly, in this data-rich age it is easier to crawl over something looking for what might be errors of fact than it is to engage with the actual argument being made.

Finally, we should remember that facts are selected. We don't use all the facts in this game, just the facts that suit our argument (and, more to the point, the other person's errors that suit our argument). The result is unedifying, often rude and seldom gets us closer to the basis for disagreement. It may well be the case that poverty has risen in the USA. Or indeed it may not be the case that this is so. What I know is that we can engage in a spat about facts without getting any closer to answering the actual question!

And Macclesfield is in Cheshire!

....

Dear doctors, if you're going to be fussbuckets can you at least get the facts right

****

The media is all afroth with the latest press release from "The Doctors" about us getting a little chubby. This is accompanied by images of podgy middles being manhandled and reams of self-righteous judging of other folks' lifestyles.

In a report spelling out the problem in stark terms, the academy says doctors are "united in seeing the epidemic of obesity as the greatest public health crisis facing the UK.

I mind greatly about doctors wanting new taxes, controls and so forth - nannying fussbucketry is not just wrong it's immoral. But I mind even more about the facts - and these doctors can't even be bothered to get these right.

Firstly there's the epidemic - well there isn't one. This is an epidemic:

A widespread occurrence of an infectious disease in a community at a particular time.

These are doctors for heaven's sake, surely they care just a little about the precision in using medical terms? Or am I crediting the fussbuckets with giving a damn? Obesity isn't infectious - I'm not going to catch it from the fat person round the corner however hard I try. I'm going to get fat if I eat too much food and exercise too little - nothing catching at all.

Then there's the line that obesity is "getting worse". This too is wrong - here's the Joseph Rowntree Foundation (not a regular enemy of nannying fussbuckets) on the subject:

Over time, there is little sign of the inexorable rise in obesity that underlies some of the concern about the issue. Rates for children did rise and peak in 2004 but have since fallen and are now no different to what they were in the late 1990s.

So we're fatter than we were in the 1980s but not actually getting even fatter.

So those doctors have misused (quite deliberately) the term 'epidemic' and lied about obesity getting worse. They also ignore the pesky research showing that being a bit chubby isn't a problem - it might even be good for us:

Results showed that overweight people were six per cent less likely to die during the average study period. Mildly obese people had a five per cent lower rate of premature death.

The research showed that men and women hold an equal advantage when it comes to being overweight, and that the conclusion was not affected by age, smoking status, or region. 

Has anyone considered that what we call "normal weight" is, in truth slightly underweight?

Even if we accept the (plainly wrong) we're all getting fatter argument, the nannying fussbuckets are after the wrong targets - fizzy drinks and burgers:

Over the last 10 years, the consumption of soft drinks containing added sugar has fallen by 9%, while the incidence of obesity has been increasing. And 61% of soft drinks now contain no added sugar. 

Doesn't look like full fat coke's to blame - the burgers then?

 ...there was no significant association between increasing takeaway and fast food consumption and obesity as measured by BMI corrected for age and gender. This is not a new finding. For example, French and colleagues found no significant relationship between frequent consumption of fast food and being overweight in their analysis of a cohort of 11-18-year-old boys and girls. Similarly, Simmons et al found no correlation between increasing takeaway consumption and obesity measured by either BMI or waist circumference.

Blast - it's not the burgers either.

This is the problem - there is a problem but the doctors are too fussed about nannying us to ask why and what might be done. Here's our friends at JRF again:

There is also a difference for adults with a clear social gradient for women (with 31% of the poorest being obese compared with 19% of the richest) but pretty much equal rates for men across all incomes (between 24-29%).

Spotted that - I'm sure you have on your high street. Let's call it the Jack Sprat Principle. But why? Why are working class women so much more likely to be obese - trying to understand this might be a whole load more valuable than banning advertising or taxing fizzy drinks.

This is the problem with public health - too terrified to target risky groups, our fussbuckets find it simpler to propose whole population solutions to problems that, in truth, affect only a minority of that population.

However, the biggest lie today is the idea that the solution lies in ever more draconian controls, more taxes and taxpayers money poured into unpleasant and aggressive "public health" campaigns. And all because - we're told without evidence or foundation - that slightly less than 3% of NHS resources goes to deal with the health problems contingent on obesity.

....

Poor fact-checking and shoddy research - an everyday story of the BBC's health reporting

****

Others have posted about the BBC's embarrassment at Panorama's dodgy statistics on the impact of minimum pricing:

It's quite incredible that Panorama's fact-checkers are so shoddy that they didn't see that claiming 50,000 old people dying is quite laughable when their source has previously declared - experts as they are - only 3,060 in the entire population!

There are two aspects of this story that I find disturbing and a third that begs a question about the value of retractions.

The first aspect is the utter credulity of BBC researchers and programme-makers when faced with information from "academic" sources. The 50,000 figure was so obviously inaccurate yet no-one on the Panorama team saw fit to ask for it checking. Indeed it became the hook around which the scare-mongering story was constructed. Even to the point of getting one of their favourites - Joan Bakewell - to front up the delivery of this punchline. That this figure is so completely inaccurate beggars belief but its ready acceptance by Panorama's producers underlines the core problem with public health reporting - the lies of the campaigners are almost always taken without question, without any challenge.

The second disturbing aspect of the story is the way in which researchers - in this case at Sheffield University - circumvent any peer challenge. Indeed, when a researcher from Panorama rang them with the programme idea the School of Health and Related Research at the University of Sheffield jumped at the opportunity to push their work by appearing on a prime time TV documentary. The "research" commissioned by Panorama reaches the ears and eyes of the SoHRR team's audience - public health professionals, political decision-makers and the concerned public - without the pesky process of peer review. The BBC may now have corrected the error but SoHRR remain unchallenged as the leading proponent of minimum pricing for alcohol as a public health tool.

Wherever we look the public health fanatics appear to be more-or-less making stuff up - third-hand tobacco smoke, fizzy drinks making you fat, e-cigarettes being carcinogenic and minimum pricing for alcohol saving lives while not affecting the drinking pleasure of anyone. As Chris Snowden puts it:

It is garbage. Not just the usual 'this sounds like a bad guess' garbage, but full-on 'no amount of squirming can get us out of this, we're going to have to retract it and re-edit the programme' garbage. The figure is more than four times larger than their dubious methodology can allow—a difference of some 38,500 lives—and thanks to the efforts of one of Dick Puddlecote's readers, the Sheffield University team have confessed to "human error". Take this with as much salt as you like, but they claim that someone accidentally put the wrong figures into the computer when Panorama commissioned the research. Truly a case of garbage in, garbage out.

As to that human error - it strikes me that this is poppycock. The SoHRR team were so cock-a-hoop at the Panorama request that they bunged out the more sensational, headline-grabbing research without bothering to test, check, cross-reference or apply the most basic of common-sense. And they did this knowing that Panorama would swallow any set of spurious "facts" they sent through. Perhaps we need to get in touch with Sheffield University's grandees and ask them to look into this "error". Although, if the University of Sterling are a guide, the reaction will be to use every means to prevent full public disclosure.

Finally we should ask what gain will come from the BBC's correction? Hundreds of thousands watched a Panorama documentary founded on a lie. A few hundred of those people will see the retraction or will watch the new version of Joan Bakewell's endpiece. Add to this those who saw the reference on the BBC main news, read articles in the Daily Mail, Guardian and Daily Telegraph and we can see that the BBC's quiet little correction represents a victory for the liars in our public health industry. I fully expect variations of the 50,000 figure to crop up in reports, documents, advice leaflets and a host of other sources. And for the medical mafia to roll it out verbally - without any challenge - again and again.

....

Sorry "Bureau of Investigative Journalism" we are not getting fatter...

****

Some people calling themselves the "Bureau of Investigative Journalism" have fallen at the first hurdle:

Britain is getting fatter.

This is not true.

...the latest figures show that obesity amongst men has fallen to 22% and the female obesity rate has fallen to 24%.

Now let's be clear that there was a dramatic increase in levels of obesity but it is now falling. On almost every measure, rates are falling and have been for at least three years:

The rapid rise in child obesity may be levelling off, according to figures. Experts looked at data forecasting levels of childhood obesity to 2020. Analysis indicated a 17 per cent drop in the forecast number of overweight girls aged two to 11 and a 4 per cent drop in the expected numbers of obese girls of the same age. There was a 5 per cent drop in the forecast number of overweight boys, and a 7 per cent drop in the forecast numbers of obese boys.

And you call yourself investigative journalists - can't you even check a few simple facts?

 ....

Things that aren't true...

****

...include this tweet:

We spend less on protecting the natural environment than Chelsea spends on their footballers. #greengrowth

Now Chelsea spend a lot on money on wages - around £172 million - but I'm prepared to bet that "we" spend a good deal more than that on protecting the natural environment. Here's some indications:

• The budget for Britain's 15 national parks totals around £70 million
• Every local planning authority provides an ecology and arboriculture planning service
• Millions are spent managing publicly owned open country outside national Parks (e.g. Ilkley Moor)
• Natural England's budget is £155m and the Environment agency spends £652m
• The Forestry Commision has a reported expenditure of over £49m

That's around £1 billion spent protecting the environment by various arms of government. And I haven't included the money spent by coastal authorities, the Ministry of Defence, the Department of Energy & Climate Change and the Department of Communities & Local Government.

But there's more - the water industry invests around £80 million each week and a fair chunk of this is spent on improvements to the natural environment (some would say at the expense of the distribution network). And then there's the millions spent by farmers, landowners, estate managers and other businesses on maintaining and protecting land. Plus of course, the National Trust, Woodland Trust, assorted Wildlife Trusts and conservation organisations.

And to finish there's the millions we all spend on our gardens, on bird tables, on trees and on - you know - protecting the natural environment.

Despite the tweet being wholly incorrect, it will do the rounds and people will believe it. The statement will crop up unchallenged in the press and on the BBC. This is how lies win.

....

Obesity rates are falling not rising - so why must doctors lie about this?

****

We've got rather used to the various clubs of doctors - BMA, RCS and suchlike - telling us that alcohol consumption is rising when it isn't (even the BBC now recognise that it is for heaven's sake). Now these New Puritan institutions are lying about obesity:

According to the latest research, 48% of men and 43% of women in the UK will be obese by 2030, a trend that will significantly increase the prevalence of strokes, heart disease and cancer, and lead to higher costs for the NHS.

Now unfortunately, the Guardian doesn't link to that latest research. However, the latest actual statistics from the Office of National Statistics tell us that rates of obesity are falling:

Despite the government ignoring the anti-obesity lobby's urgent suggestions for traffic light labelling on food and suchlike, the latest figures show that obesity amongst men has fallen to 22% and the female obesity rate has fallen to 24%.

So something must have changed - either the Academy of Medical Royal Colleges are scandalously misinformed or else the British public has been on a massive eating binge in the last year (those latest figures are from 2011).

Ah but what about the children say the doctors:

Charlie Powell, campaigns director of the Children's Food Campaign, applauded the academy's intervention. He said: "Andrew Lansley should act on this excellent set of robust recommendations, but his track record suggests that he will once again ignore the advice of our best medical experts."

Ah yes those children - I don't have figures for everywhere but I'm pretty sure Bradford won't be out of the ordinary in terms of obese children -  here, for Year 6 children, there has been a slight fall over the past three years and the rate remains below 20% (marginally above the national average). And note that this uses the very discredited measure of Body Mass Index.

None of these actual facts supports the argument of the Academy that we should set about banning advertising & sports sponsorship, fast food outlets near schools and celebrity endorsement. Let alone introducing "fat taxes".

There really isn't a growing obesity crisis and doctors should help people who get too fat rather than seek to punish the vast majority who never get past being slightly chubby.

....