Some Tories say voters are tired and weary of austerity*. The voters are right.

Let's start with the sums. There's a simple equation to tell us what primary budget balance we need to stabilize the debt-GDP ratio (the primary balance is annual borrowing excluding interest payments). The equation is:

d * [^(r-g)/_(1+g)]

where d is the debt-GDP ratio, r the real interest rate and g is trend growth. D is now 86.5% and r (measured by the 20 year index-linked gilt yield) is minus 1.7%. Let's be cautious and call g 1.5%. Our equation then tells us that we can stabilize the debt-GDP ratio even with a deficit of 2.7% of GDP. However, the present deficit is only 0.9% of GDP, and the government's current plans envisage a surplus after 2019.

The government can therefore borrow a lot more than it is currently and still reduce the debt-GDP ratio. That's the beauty of negative real interest rates.

This poses the question: should we really aim to cut the debt-GDP ratio as fast as the Tories plan?

There are two reasons to think not. One is that fiscal tightness depresses GDP growth and this cannot be adequately offset by monetary policy when we're near the zero bound. The other is that there is a global (pdf) safe asset shortage, which means that high stocks of government debt (safe assets) are now more sustainable than they were in the 80s or 90s.

There are two counter-arguments to this. One is that we need to cut government debt now to give us fiscal space for a loosening in the next recession. The other is that the strong demand for government debt might not last. These arguments, however, seem to me to be mutually exclusive: in a downturn, demand for safe assets will rise**.

There's another reason to end austerity. It's that Osborne's way of doing so achieved the impressive physical feat of being half-arsed and cack-handed. Serious efforts to cut government spending would have done three things:

- They would have created an ideological climate conducive to smaller government.

- They would have reduced the functions of government: it's easier to do less with less than to do more with less. For example, if you cut police numbers, you should cut the crimes they investigate - say by legalizing drugs or sex work. If want fewer prison officers you should cut the number of prisoners. And if you want a slimline civil service, you shouldn't give it the gargantuan task of organizing and negotiating Brexit.

- They would empower workers who know the ground truth of how the public sector wastes money.

The Tories have done none of these. As Sam says, they "have totally failed to make a broad-brush case for free markets." They've tried instead to cut spending by top-down managerialism. The upshot has been that those cuts have been unsustainable. As Rick says, even small further cuts "might be enough to tip some public services over the edge."

You might reply here that government spending, at just under 40% of GDP now, is too high given that an ageing population will increase such spending in future.

In this context, my chart tells an important story. It shows that there's a significant negative correlation between five-year changes in the share of government spending in GDP and GDP growth. Faster GDP growth - such as in the early 70s, late 80s and early 00s - is associated with falls in public spending as a share of GDP. And slower growth - such as the recessions of the mid-70s, early 80s and 2008-09 - see increases in that share.

If you want to shrink the ratio of public spending to GDP, therefore, it's much easier to do so if the denominator is rising strongly. We must, therefore, try to increase trend growth. In this context, it might be that austerity is counter-productive in another sense - because, as Simon says, it might have depressed trend growth. Whatever, policies to increase productivity are imperative.

You might object here that such policies won't work because it's just damned hard to increase (pdf) trend growth. Maybe. To which I say: let's test the hypothesis with rigorous policy action.

* I'm defining austerity by the change in the cyclically-adjusted primary budget deficit. This has been cut fro, 6.4% of GDP in 2009-10 to 0.9% now.

** Also, even if you plug a plausibly higher r into my equation, you still get a similar outcome - that we don't need a surplus to cut the debt-GDP ratio.