"Never mind the public finances. Look after the economy and the deficit will look after itself." For years, critics of George Osborne such as me echoed Keynes' old line. At last, the Tories are, by their own lights, heeding this call. And financial markets hate it. Does this mean the Keynesian advice was wrong all along?

No. There's a big difference between the 2010s and now.

This difference is not that Kwarteng is mistaken to think that tax cuts will stimulate long-term productivity growth. He probably is, but this isn't the main problem. Even if he had announced the best supply-side policies imaginable, these would not have a significant impact on the economy in the next few months simply because it takes time for capital and labour to learn of new and better opportunities and move to them. As Banerjee and Duflo showed in Good Economics for Hard Times, economies are sticky and slow to adjust to change.

Instead, we are closer to full employment now than we were ten years ago. And full employment changes how macro policy should be conducted. When there is unemployment any fiscal policy that boosts aggregate demand will lead to the greater utilization of unemployed assets (capital and labour) and hence to increased real output. At full employment, however, this cannot happen and so we get inflation instead. Tax cuts might have boosted growth ten years ago when there were idle resources, but now we are nearer to full employment they are more likely to lead to higher inflation.

Which is what financial markets expect. Their fear of higher inflation has driven five-year gilt yields up to 4.4%, compared to only 1.4% as recently as May*.

This raises three issues.

One is: how much will this fiscal stimulus add to demand? In this context, the fact that the tax cuts are skewed to the rich is a good thing: the rich have a high propensity to save and so won't boost aggregate demand and inflation much. But even so, the impact is positive - as of course is the impact of capping energy prices**.

A second issue is: how close are we actually to full employment? The official jobless level is just over 1.2m. That's less than the number of vacancies, and its lowest since 1975. On top of this, though, there are 1.7m people out of the labour force who would like a job. Adding these to the officially unemployed gives us a level of 2.9m, which is the lowest since data began in 1992. Official figures show that there are also 2.8m people who are working fewer hours than they'd like - although this is the lowest number since before the financial crisis.

Whether such numbers mean we are actually at full employment is moot: there's always some frictional unemployment and under-employment simply because of mismatches between supply and demand***. What we can say, though, is that we are much closer to full employment than we were ten years ago. Back then there were one million more under-employed and two million more unemployed and inactive wanting a job. Which means it is plausible that a given boost to demand now will lead to a worse output-inflation mix than it would have done in 2012.

Thirdly, even if we are at full employment, is this inflationary? In the past companies have responded to high demand by figuring out (pdf) ways to raise productivity. That's one reason why full (male!) employment in the 50s was not inflationary. But markets can be forgiven for doubting whether a repeat is possible. For one thing, forty years of mass unemployment has made bosses lazy: in relying upon abundant labour they've lost the habit of thinking how to raise productivity. For another, high policy uncertainty deters companies from investing in the capital equipment that might raise productivity. And for yet another, high demand now is less likely to lead to a big influx of migrant labour, in part because sterling wages no longer buy so much in foreign currency.

It's likely, therefore, that a fiscal stimulus today will be more inflationary and less good for output than it would have been ten years ago. Hence the market reaction.

It's in this context that Sunak's health and social care levy made sense. if we want more health and social care the question is not: where will the money come from? It can be borrowed. Instead, it is: where will the real resources come from - the carers, nurses and premises for care homes?

When there is mass unemployment, the answer's simple: they come from the dole queue and vacant property. At full employment, however, the answer is different: they must come by shifting people from one employment to another. The logic for higher national insurance contributions was that it would do this: it would depress consumer spending and hence facilitate a shift in labour from retail and hospitality to the care sector.

Granted, the implementation of the policy should have been better, for example in helping people more to retrain as health and care workers. But the macroeconomic logic was sound, assuming we were near full employment. Today's Tories are rejecting this logic. That wouldn't be a problem if there was significant unemployment. But there might not be, and indeed Kwarteng's attempts to cajole UC claimants into working more suggests that he believes there is not.

Now, you might question the facts here. Maybe we are still some way from full employment, or maybe our nearness to it won't be as inflationary as feared.

For my purposes, however, this is not the point. The point is that the policy that's right for unemployment is not the same as that which is right for full employment. Macroeconomic economic policy should be a merely empirical matter, not a matter of ideology or sloganeering. In this respect, however, the Tories have been consistent: they were wrong in the 2010s, and might well be wrong in the 2020s. But that's the thing about them: they are so much more ideological than we Marxists.

* Granted, all the rise is due to a rise in real yields: Bank data shows that inflation expectations have actually come down since then. But that is because markets expect rate rises to curb inflation: higher mortgage rates will crater discretionary consumer spending.

** Yes, the price cap cuts inflation as a mathematical fact. But it raises it as a behavioural one: because we'll have more money to spend than we otherwise would, demand will be higher which is potentially inflationary.

*** From a macroeconomic point of view, "full employment" is defined as the unemployment rate at which inflation takes off. This need not be the same as full employment from a human point of view.