Richard Murphy says we can afford to give public sector workers a pay rise. I think I agree, for a reason he doesn't mention.

It's trivially obvious that Theresa May was wrong to claim that there's no magic money tree. In fact, in theory, there are several.

One, as Richard says, is the multiplier effect: higher public spending brings in higher tax revenues, both directly as public sector workers pay more tax and indirectly as their extra spending generates more economic activity.

Another is that governments can print money to pay for extra borrowing. The Bank of England has done £435bn of this.

A third is that government borrowing costs are negative in real terms. Despite a slight rise recently, 20 year index-linked gilts yield minus 1.6 per cent. This means that for every £100 the government borrows, it will have to pay back only £72. This fact, combined with even moderate economic growth, means the government could borrow more and still see government debt shrink over time relative to GDP. It is just plain wrong, therefore, to claim that paying public sector workers more will impose a burden upon future generations.

There are, therefore, magic money trees. The fact that Ms May could claim otherwise is yet more evidence of the stupidity of her government. And the fact she wasn't sufficiently corrected for her error is evidence of the inadequacy of the media.

Just because you've got a magic money tree, however, doesn't mean you must shake it. There's a simple reason why governments have been loath to use these trees: inflation. Higher government spending means increased aggregate demand, which - it has been thought - leads to higher inflation.

It's inflation, and not the lack of a magic money tree, which has traditionally been the obstacle to higher public spending.

Which brings me to why I agree with Richard. This danger isn't especially great.

I don't say this because I believe a little more inflation will be a good thing. I suspect that higher inflation might actually be a contractionary force as it might encourage households to save more - a danger which is especially great given that the savings ratio is at a record low now. I share Eric's concern that lower real rates now might actually be a bad thing.

Instead, I've another reason not to worry much about inflation. It's that additional economic activity doesn't seem to be as inflationary as previously thought. We know this because wage inflation has stayed low despite unemployment falling to a 42-year low. The UK's rising inflation now seems to be just the effect of sterling's devaluation, which should be one-off. (This isn't a quirk of the UK economy: much the same is true in the US too).

This might be because there's hidden unemployment: the Resolution Foundation estimates that underemployment is greater now than in the mid-00s. Or it might be because, as Andy Haldane has said (pdf), changing working practices have led to a flatter Phillips curve. As Simon says, the Nairu has fallen. Or it might just be that in an open economy the trade-off between growth and inflation isn't as great as people think.

Whatever the reason, the inference is the same. The government might be able to borrow more to pay public sector workers more simply because the inflation constraint on it doing so is weak.

Of course, I might be wrong here. Maybe there will come a point when stronger economic activity is inflationary. Or maybe higher public sector wages will lead to general inflation if it triggers higher private sector wages. This is an especial risk because wage rises won't be offset by productivity gains.

If I am wrong, though, there's a simple solution: higher interest rates. A mix of these plus looser fiscal policy would have some advantages. They'd take us away from the zero bound and so give the Bank of England more room to loosen policy in the next downturn. And they'd tend to dampen down house prices and financial speculation.

I suspect, then, that the risk of giving public sector workers a pay rise is small. Not only is there a magic money tree, therefore, but we should shake it.