The cost disease

Shortly after reading William Baumol’s new book ‘The Cost Disease: Why Computers Get Cheaper and Health Care Doesn’t‘, I thought that its subject matter – rising healthcare costs – may make a good subject for a post. However, John Appleby beat me to it with this excellent ‘Data Briefing’ in the BMJ. In this piece, Prof. Appleby discusses how the inexorable rise in healthcare spending will consume an ever larger portion of GDP. The cost disease does get a mention albeit only briefly:

“the prices of the inputs to healthcare have tended to rise in line with, or even faster than, costs in the economy as a whole—a reflection of the “cost disease” identified by William Baumol in labour intensive industries where the productivity increases that could offset rising pay costs are hard to achieve.”

For me, this is one of the most salient points in the whole discussion, and as such warrants further exploration. And, as Baumol reassures us, the cost disease analysis provides us with good news on the whole since the nature of the cost disease means that we will still be able to afford healthcare as time goes on.

The idea of the cost disease is a simple one. Overall the economy experiences productivity growth; innovation means that the marginal product of an hour’s work increases. But there are industries that experience productivity growth greater than the average and some that experience it less than the average. What’s more, it is almost always the same industries that lie on each side. Baumol and colleagues call the less than average group the ‘stagnant’ sector. These stagnant services typically involve large amounts of labour – a certain element of handicraft. These include healthcare, education, performing arts, and so forth. But why then do their costs rise?

In the non-stagnant sector increases in productivity (should) lead to increases in wages. A worker on an automobile production line where productivity has increased by 4% should expect to see wages increased by 4%. This means that the manufacturer’s profits remain the same and the price of the car remains the same even though the worker’s wages are higher. In the stagnant sector wages will increase to keep in line with wage increases in other areas; but a 4% rise in wages in a firm with less than 4% productivity growth must either lead to an increase in price or a decrease in profits. It is usually the former.

It is as simple as that. Healthcare costs are rising because productivity growth in the healthcare sector is slower than average but wages increase in line with the rest of the economy. If economic growth stops, healthcare costs stop increasing. So we shouldn’t be worried. And it is because of productivity growth that we shouldn’t be worried. For our automobile worker, his wage increases relative to the price of the car each year. Goods from the non-stagnant sector become more and more affordable. Productivity growth means GDP growth so that each individual’s spending power increases year on year. So while stagnant sector services like health care take up a greater and greater proportion of the pie, that pie is growing at least as fast in size. Prof. Appleby notes in his post that healthcare spending is projected to take up 17% of GDP by 2062 compared to just over 6% today. But, GDP is projected to be three times larger by 2062.

In Baumol’s book he points to a number of examples of increasing costs in stagnant sector services. In particular, he focuses on rising healthcare costs and university tuition but also notes other areas such as funeral services and music performance.

Baumol makes a number of important points about the cost disease that are important when considering health care:

  1. The cost disease disproportionately affects the poor – even though increasing productivity should help alleviate poverty by providing more goods and services to the poor, the increasing costs in healthcare could serve as a barrier to these services.
  2. Misunderstanding of the cost disease may lead to government intervention that could make the situation worse – the frightening prospect of healthcare spending inexorably taking over a greater and greater share of GDP could lead to ill-advised interventions. In particular, governments may implement cost controls to keep cost increases below economic growth. But, this will just lead to a reduction in services without actually reducing costs. As we have seen, the cost disease is unavoidable and is a consequence of growth. In another of Prof. Appleby’s BMJ pieces, he provides a graph demonstrating the frankly ludicrous productivity targets the government has imposed on the NHS. In light of our understanding of the cost disease these targets are obviously unobtainable while maintaining the same level of growth in the economy as a whole.
  3. The private sector is liable to the same issues – since many of the stagnant sector services are public sector services, governments are liable to privatise them to control costs. But, the private sector is not immune from the cost disease and can make the same mistakes.

It is important to note that this doesn’t mean that we can’t improve services and reduce spending somewhat now. There are inefficiencies that can be eliminated. Consider the US, which spends 16% of GDP of healthcare, and how it compares to the UK, which spends 6% of its GDP, and yet there are little differences in public health outcomes. There are certainly savings to be made in the US. But, whatever savings are made, healthcare costs will still increase faster than average and will consume an ever greater share of GDP.

Overall, I would recommend Baumol’s new book. It is somewhat over simplified and may not satisfy the economist who reads it. But, it provides a number of clear and relevant examples of the cost disease that should stimulate further discussion and analysis.

The economics of the ‘nudge’: Why the UK government’s new public health policy won’t work

The UK Government recently announced its plans for “Public Health England”, with ‘nudging’ high on the agenda. What the government considers a ‘nudge’ is unclear, though it seems to mean giving people the opportunity to improve their own health. Here’s why I think this is a flawed policy.

As a recent BMJ article pointed out, there appear to be many more ‘nudges’ in the opposite direction. As long as private companies are allowed to push their delicious fatty foods and cigarettes with all the marketing that they desire, it seems that any government ‘nudge’ will be overshadowed. In the UK, the total value of tobacco sales alone was £11.3 billion, while the government suggests a figure of just £4 billion to be ring-fenced for the new public health service. True, Public Health England will be able to spend almost all of this money on ‘nudging’, but with Diageo (the world’s biggest spirit maker) alone pulling in revenue of £7.1 billion and profits of £1.6 billion in 2010/11, it’s hard to see how a measly £4 billion could counteract the industries’ marketing ‘nudges’.

This doesn’t even take in to account the utility individuals gain from scoffing that 3rd bag of Walkers, or guzzling that penultimate pint of cheap lager at last orders. Nor does it take in to account the fact that you really can’t be bothered using that new-cycle-path-shaped ‘nudge’ to get to work tomorrow, or the fact that you’ll have to sacrifice an hour of your valuable leisure time to take your kid to that new playground. Nor does it take in to account the fact that you don’t even have a bike, and you’re too drunk to even notice the ‘nudge’ on that second bottle that says “the average British drinker drinks one glass of wine a night”.

To be fair, I know very little about behavioural economics, and I have not read “Nudge” (though I will). However, my understanding of compensating and equivalent variations tells me that people are going to need some hard-cash-value incentives or disincentives to have them change their behaviour; either that or regulatory restrictions.

So, where does the ‘nudge’ stand in relation to health economics? Is it something that we should be harnessing? Will anybody be trying to evaluate the QALY loss or gain from particular ‘nudges’? Surely we should, otherwise we might be getting ‘nudged’ into accepting a policy that currently has almost no empirical support.