Slim truth in fat figures

Michael West
November 23, 2011 – 11:39AM
The Age

A leading actuary has lampooned health lobby figures on the costs of smoking and obesity as being extravagantly inflated and based on suspect methodology.
“The numbers are all over the place,” writes Geoff Dunsford in the September edition of Actuary Australia. And they are “big numbers” – the implication being that they are too big.

“Obesity costs $58.2 billion,” he exclaims, “that’s around twice the cost of age pensions!”
The sheer size of the numbers, argues the Sydney actuary, perverts government policy. It can lead to poor spending decisions. The credibility of the numbers from the health lobby is therefore critical to government policy.

The press and the public have been led to believe that the costs to the system are higher than they really are so the government can “justify use of taxpayers’ money on measures to reduce its prevalence and prevention”.

Dunsford looks at three public health issues: obesity, smoking and depression.
1. “….obesity …. drains the national budget each year by $58.2 billion”, (Sun Herald report, March 13, 2011).
2. “…smoking … costs our society $31.5 billion each year”, (Nicola Roxon, media release, April 7, 2011).
3. “Depression-associated disability costs the Australian economy $14.9 billion annually”, (beyondblue website)
In the first case, the newspaper story was based on an Access Economics report for Diabetes Australia titled, “The growing cost of obesity in 2008: three years on”.
Access Economics estimated the cost of obesity to Australia at $58.2 billion. And sure enough, this enormous headline number promptly bobbed in the press.
On Dunsford’s analysis, however, the figures are flawed, skewed by the “non-financial” estimates to make obesity seem a lot more costly to the taxpayer than it really is.
The costs break down as $3.9 billion for the health care system, $4.4 billion in “other” costs relating to lost work days, taxes forgone and other productivity losses.
Then there is the big one: $49.9 billion in “non-financial costs”. This relates to “burden of disease” or the personal cost of obesity. Dunsford asks, “how come this is included in a total in an announcement which appears – at least superficially – to represent real money costs?”
The “burden of disease” numbers are calculated by working out “years of life lost through disability and premature death” and Access came up with $6.35 million for the value of a statistical life (VSL) and $266,843 for the value of a statistical life year (VSLY).

Dunsford argues that it is taxpayers and consumers who will end up paying for all this statistical life.
“The elaborate details on labels of packaged food products in supermarkets are testimony to the current massive regulations supporting such details, but more are planned by Food Standards Australia NZ and the National Preventative Health Taskforce,’’ he says.

From there it would only be “a short step” to include take away food and restaurant meals and, already, in certain states of the US, it is a requirement for restaurants to display the calorific value of their meals in the same size print – “including on billboards!”
‘‘The cost of administering the regulations (to the government and the food industry, all of which will ultimately be paid by consumers) will be mind-boggling, but with a focus on the desire to reduce the $58.2 billion cost of obesity, such actions can readily be justified.”
Tobacco figures are smokin’

Geoff Dunsford is similarly wary of the costs estimates for smoking.
Assessing the anti-smoking lobby’s $31.5 billion cost figure – found in “The costs of tobacco, alcohol and illicit drugs abuse to Australian society 2004-05” by David J Collins and Helen M Lapsley – Dunsford once again shines the torch on the “non-financial” costs and “intangible costs”.
Of the $31.5 billion, some $19.5 billion are “intangible costs” – that is psychological costs of premature death borne by the smoker and others. Then there are $9.4 billion in “other financial costs” for productivity losses (smoko breaks perhaps?) and $2.2 billion in “non financial costs” such as unpaid labour costs.
In the Collins and Lapsley report there is a discount for savings to the health system from premature deaths. But this is only $700 million on the $1 billion in actual costs to the health system.

On the more nebulous costs, estimated by a “demographic approach”, the focus is on the additional number of persons who would have been alive today had there been no smoking deaths over the past 40 years.

“An estimate of 369,161 was provided to the authors by John Pollard (he had no other involvement with the report).”
To get to this $19.5 billion, the authors multiply the reduction in the population (369,161) by the value of the loss of one year’s life ($53,267), after adjustments.
As Dunford points out, this report puts a different value on life than does Access Economics. Whether Access prices obese people more highly than skinny smokers – or Collins and Lapsley believe smokers are worth less than one-third of the value of fat people – we can’t be sure from “the literature”.
Indeed each report mentions “the literature” and the large variation in assumptions included in “the literature”, although they also fail to explain, he says, why the numbers they adopted were relevant to their particular health problem.
In addition, the “value(s) of a statistical life” and the “value(s) of a statistical life year” adopted were significantly different ($6.35 million and $266,843 for obese people, and $2 million and $53,267 for smokers).

Presumably – and these are our words not Dunsford’s – the pricing of a statistical life would also become more complicated when calculating the demographic of people who are both obese and smokers. Do we just average out the $6.35 million and the $2 million?
Dunsford however does point out the gross hypocrisy in the government’s position on smoking and revenue. Governments reap very fat profits from smokers.
Subtracting the financial costs of smoking to the health system at $300 million, plus taxes forgone (from statistical smokers) at $2.9 billion, from the $6.7 billion in taxes levied by state and federal governments on tobacco products, leaves $3.5 billion in profit.
Dunford says the higher welfare payments to smokers could be offset by the pension savings from higher smoker mortality. Still, a $3.5 billion profit from smokers is a tidy amount for the budget.

When it came to the publicity for the “plain packaging” initiative, it would have been helpful, says Dunstan, to cite the $31.5 billion in “costs to society” rather than a more realistic figure.

“Indeed, assuming the media release’s (Roxon’s office) expected reduction in adult smoking from the current 16 per cent of the population to 10 per cent is achieved, the reader could be forgiven for estimating the ‘cost’ to fall by … $11.8 billion,” says Dunstan.
But such an assumption would be wrong as the methodology is flawed.

“The problem with the … definition of costs is the way in which past drug abuse is incorporated into the costs for a given year,” says Dunsford. ‘‘Indeed, if all smoking stopped, (this) methodology would still generate a large cost of smoking in the next year by virtue of the effect of the past deaths.
“This is rather counter intuitive! Arguably it renders the methodology meaningless for the purpose of addressing cost reduction initiatives”.
‘Burden of disease’ missing
Finally, Dunsford points out that in the case of the beyondblue calculations, the $14.9 billion of annual costs to society from depression did not include a ‘’burden of disease’’ number.

“Does this reflect the often suggested lack of interest by the government in mental health problems? Not so. Burden of disease numbers are available which show DALYs (disability adjusted life years) due to depression are significant – particularly when associated causes of death, like suicide, are included.” he says.
He estimates that about $33 billion of non-financial costs could be added to the annual cost numbers for the personal impact on the loss of wellbeing from the burden of depression.

Dunsford’s work is further proof we can’t place much store in lobby group costs claims. It’s more a case of plucking out a big number and working out some methodology to justify it.

Read more: