In this submission to the Commons Treasury Select Committee tomorrow, Sean Boyle and John Appleby of the King's Fund explain why Tony Blair is likely to fail to reach his target of increasing Britain's spending on health to European Union levels:

Tony Blair's commitment to raise total healthcare spending in the UK to the average of the European Union by 2005-06 has left unanswered how much it would cost. Perhaps for obvious reasons – it is going to cost a lot – the Government has avoided clarifying its apparent spending target.

After Mr Blair first made the pledge last year, it soon became clear that the Government's notion of what the EU "average" meant left something to be desired. Simply adding up the percentage healthcare spends in all EU countries and dividing by 15 was not, most commentators agreed, the right way to do it. For the Government it gave a target on the low side – 8 per cent of GDP.

But a more correct average -- the weighted mean -- should have been used. This sums the absolute levels of healthcare spending and GDPs across the EU and divides the former by the later.

And instead of calculating this average for all EU countries, the UK should have been excluded, as inclusion reduces the average (because UK spending is currently below the average), but will move the average up as spending in the UK increases. A weighted mean that excludes the UK produces an EU average of just over 9 per cent.

But this new average is still wrong as it is based on spending reported by the OECD for 1998. By 2006 – the target date – the EU average will have changed, almost certainly being higher than in 1998. Based on trends in the EU average spend over the past 40 years, we have calculated that the weighted EU average healthcare spend as a percentage of GDP (and excluding the UK) will reach 10.7 per cent by 2005-06.

Meeting the Government's 8 per cent target by 2005-06 is likely to require an additional £14.2bn over the three years 2003-04 to 2005-06. This is a cash increase of more than 22 per cent. In real terms this is equivalent to a rise of 4.3 per cent a year, which compares with an average real increase over the past three years of just over 7 cent.

However, to meet the 1998 weighted average – excluding the UK – (9.03 per cent) would require a £26bn cash increase (over 40 per cent more than next year's allocation), equivalent to real annual growth of just under 9.4 per cent a year for three years.

Finally, to meet the EU weighted average (excluding the UK) as estimated for 2005-06 will require cash increases of £45bn – 70 per cent more than next year's allocation, equivalent to real annual growth of nearly 17 per cent.

So the significance of a government commitment to meet the EU average depends on its precise definition and on how spending on health care in the other EU countries changes over the next few years.

At one end of the scale, on the basis of the Government's definition, there should be little difficulty in meeting this target. Although the real rise is more than 4 per cent a year, which is greater than the predicted real growth for GDP, it is still considerably less than has been allocated to the NHS over the previous three years and should have no crucial im- plications for taxation policy.

At the other end of the scale, using more accepted measures of the average, very large sums of money will be required if the UK is to meet the EU average total health care spend by 2005-06 – up to £45bn. Such sums raise questions about the balance of spending across government departments in general and the possible need to raise taxation levels. The choice about the level of NHS spending will, of course, remain a political decision.

Sean Boyle is visiting fellow, health and social care, at the London School of Economics. John Appleby is director of health systems at the King's Fund.