It is quite important to know that virtually every day since 1948 the NHS has been said to be in crisis,” Nick Timmins says in the introduction to his magisterial history of the British welfare state. It has always been in someone’s interest to say so. Sometimes it has even been true. It may well be true today after seven years of the tightest spending settlements in its history.

Such moments bring forth inevitable calls for action. One old idea is gaining a lot of traction at present — that we should raise some form of hypothecated or earmarked tax to pay for the NHS. The argument is attractive. After all, people say that they would be willing to pay more for the NHS. But politicians are reluctant to raise taxes. The link between higher taxes and more money for, and better services from, the NHS is unclear. The pain from paying more tax is all too clear. So why not link the two explicitly? Voices as diverse as Nick Boles, the Conservative former minister, and Lord Layard, the Labour peer and economist, have been making the case in recent months.

Reports of such calls for a hypothecated tax almost always contain a caveat along the lines of “but the Treasury, not wanting to lose control of tax and spending, has traditionally been opposed to hypothecation”. The implication is that a rather reactionary and fuddy-duddy finance ministry is standing in the way of a rather good and radical idea.

Except on this one I think the Treasury view has quite a lot going for it. And its reasons for being suspicious of hypothecation are a lot less cynical than a simple desire to retain control. Because you never can, or would never want to, actually link what we spend on health to what we happen to raise from a particular tax, any serious attempt at hypothecation is almost bound to end up being little more than an exercise in deceiving the voters.

As ever, a good starting point is clarity about aims and about meanings. There are two, quite different, objectives behind calls for a hypothecated tax for the NHS. The first is to relieve immediate pressures. The second is to create a stream of revenue that can grow over time as demographic, technological and other changes inevitably increase the amount we need to spend to provide an acceptable level of healthcare.

People also mean rather different things when they talk about a hypothecated tax. As John Appleby, veteran chief economist at the Nuffield Trust puts it, hypothecation is a bit like Brexit — “hypothecation means hypothecation” rather as “Brexit means Brexit”. It has a tendency to mean whatever its proponents want it to mean.

One version is simply to announce that some tax rate or another will be increased and that the additional revenue will be spent on the NHS. That is designed to meet the first objective, to relieve immediate pressures, and is exactly what Gordon Brown did back in 2002 when he raised national insurance contributions in order, he said, to increase NHS spending. As a political stratagem, that was effective. But equally he could have linked the increase to the expansion of the tax credit programme announced at the same time. There was no actual link between the tax rise and the increase in NHS spending.

In any case, NHS spending soon rose by vastly more than the additional revenue raised by this one tax increase. The amount we now spend on the NHS is clearly wholly independent of the 2002 rise in national insurance contributions. In the (not very) long run, a tax increase like that is just like any other tax increase.

I have no great problem with this as a short-run political expedient, but that’s all it is and it is wrong to suggest it is any more than that.

Hard hypothecation (just like Brexit, it comes in harder and softer varieties) would tie NHS spending to revenue from a particular tax. That tax could be raised over time to meet growing needs. One suggestion is that a modified version of national insurance contributions should pay for the NHS in its entirety. We would decide how much health spending we wanted at the start of a parliament and adjust rates to achieve that level of spending. Though why we would want NI as opposed to any other tax to rise over time in line with increased health spending is a bit of a mystery.

Other models are available, but all suffer from similar problems. There is no world in which we would actually want to tie NHS spending to what we raise from a particular tax. Look at what happened a decade ago. In 2007-08 national insurance contributions brought in £100 billion, while we spent £102 billion on the NHS. Three years later, after the financial crisis, national insurance contributions were bringing in £96 billion while NHS spending had risen to £121 billion. In a world of hard hypothecation, NHS spending would have been cut or national insurance contribution rates raised during a recession. Either policy would have been crackers. Proponents of hypothecation say that you could build up a fund and borrow from it during recessions and pay back in booms. Far from introducing the clear and transparent link between revenue and spending, this would increase complexity and opacity.

We almost certainly are going to have to increase the fraction of national income we devote to health in future. That will mean higher taxes. Governments will need to explain that honestly. Attempts at hypothecation are more likely to get in the way of that explanation than to help it.

This article was first published in the Times newspaper and is reproduced with permission.