Sarah Wollaston is wrong on so many levels

Almost unbelievably, the Tories are planning to ban retailers from offering multi-buy discounts on half the food sold in Britain, amongst other ridiculous measures. As I argued this week, these policies are essentially plucked out of the air with no thought given to whether they might work or what damage they will cause.

Sarah Wollaston has hit back at critics, saying...

To those who say this is ‘nanny state’ stuff, perhaps if they don’t care about the personal lifetime cost for children, the costs of obesity related ill health are now estimated to be greater than our spend on police, fire service & judicial system combined.

— Sarah Wollaston MP (@sarahwollaston) May 30, 2018

This is bad logic, bad economics and bad philosophy.

For a start, there is no reason to compare these particular costs, and such a comparison does not imply that any particular course of action should be taken. The same figures could be equally be cited as evidence that we should spend more on the police.

Secondly, the figures are simply wrong. The UK spends £15.5 billion on police and fire services whereas the gross cost of obesity to the health service is estimated at £6 billion. Wollaston is almost certainly referencing a discredited McKinsey report which lumps together all diabetes spending on top of all obesity spending. As Mark Tovey pointed out in an IEA report...

In 2014, a 120-page report called ‘How the world could better fight obesity’ was released by the McKinsey Global Institute. The authors were promoting to the world’s governments a set of 44 interventions, and in their appeal to the UK they wrote:

‘ ..the government currently spends about £6 billion a year on the direct medical costs related to being overweight or obese... It spends a further £10 billion on diabetes. The cost of obesity and diabetes to the healthcare system is equivalent to the United Kingdom’s combined ‘protection budget’ for the police and fire services, law courts and prisons; 40 percent of total spending on education; and about 35 percent of the country’s defence budget’. 

Though the £6 billion and £10 billion look impressive together, especially when compared to various departmental budgets, they cannot legitimately be summed. The £6 billion is an inflation-adjusted version of a figure from a 2011 study by researchers at the University of Oxford, who included the proportion of diabetes costs attributable to overweight and obesity in their estimate. So the McKinsey report was double-counting, and also including costs wholly unrelated to body size when it added £10 billion on top. That did not stop the Telegraph , the Daily Mail , the Independent, the Guardian and even the Chief Executive of NHS England from uncritically reporting the offending figure.

Thirdly, the gross cost to the health service is irrelevant to taxpayers. It is the net cost to overall government spending that matters. Tovey estimates that the net cost of obesity to the state does not exceed £2.5 billion and is probably significantly less. As he says in the conclusion...

The burden on the taxpayer narrative has been exaggerated by anti-obesity policy wonks, looking to make their esoteric proposals newsworthy during a time of slow motion crisis in the NHS. Past researchers have completely omitted the fact that reducing body weight entails its own costs, because the extra life years gained lead to extra pension, healthcare and benefit spending by the government

Fourthly, the cost-to-the-taxpayer argument is only worth making if the policy being proposed is going to reduce that cost. Not only should it reduce the cost, but the savings must exceed the cost of the policy. Presumably even Wollaston does not believe that taxing milkshakes and banning food discounts is going eliminate the costs of obesity, so how much does she think they are going to reduce the obesity rate by? One per cent? Five per cent?

Let's say it is five per cent and that a five per cent reduction in obesity will cause a five per cent reduction in the net costs of obesity. That would be a saving to taxpayers of no more than £125 million per annum. It seems plausible that the cost to consumers of banning multi-buy discounts would exceed this, but neither Wollaston nor any other anti-obesity campaigner has bothered to do a cost-benefit analysis. They have no idea how much it will cost and don't know if it will reduce obesity at all, let alone by how much. Given that a ban on multi-buy discounts for alcohol failed to cut sales of alcohol in Scotland, there is good reason to think that it will have the same effect on obesity as every other nanny state policy that has been tried, ie. none.

Finally, Wollaston's argument fails at the most basic level of reasoning. She is trying to refute the accusation of paternalism by appealing to self-interest. It is not about coercing citizens into living a state-approved lifestyle, she says, but about saving taxpayers money. This is a common retort from nanny statists and is OK so far as it goes. The trouble is that they usually get the figures wrong (Wollaston is a repeat offender in this regard) and they don't really care about economic efficiency. Most preventive medicine, if successful, creates a net cost and smoking saves taxpayers a fortune, but this doesn't stop Wollaston trying to coerce people away from smoking.

Nevertheless, arguments about externalities can be valid. The problem is that they do not trump all other considerations. Wollaston avoids having a conversation about the normal policy-making considerations - costs, benefits, unintended consequences, liberty, efficiency, waste, etc. - by saying 'muh, taxpayers'. 'Something must be done', she says. 'This is something therefore it must be done.'

Her argument, such as it is, is no different to that of an authoritarian ruler who removes jury trial and tweets...

To those who say this is ‘police state’ stuff, the costs of crime are now estimated to be greater than our spend on road maintenance, foreign aid & defence combined

I am not equating BOGOFs with jury trial. The point is that it is a non sequitur. The mere existence of costs does not justify illiberal policies, particularly when there is no guarantee that the policies will reduce those costs and when the policies themselves will create new costs.