27 June 2018

The evidence is clear: government spending means less growth


Having agreed a £20 billion “70th birthday present” for the NHS, the Government is now, naturally enough, facing pressure to raise spending in other areas.

If Theresa May agrees to raise NHS spending and defence spending, schools will presumably be next, along with housing, transport, welfare, and so on. A few billion here, a few billion there, and pretty soon you’re talking real money. All the magnificent efforts put into cutting spending during the long years of austerity will have been for nothing, as spending zooms up again.

Now, of course, higher spending allows us to do more of lots of things governments and broader society likes — we get to help the poor, treat the sick, educate our kids and fight terrorism. We might dispute whether all of these things are best done via the state, but if we’re doing them via the state one can at least see some upside in doing more of them rather than less.

But what about the downsides? Obviously there are moral issues to consider. Taxes to pay for public spending (or to repay the debts that paid for public spending) involve confiscating people’s money, so there are ethical problems with doing more of that than is strictly necessary. But there is also the problem that raising public spending and taxation, relative to GDP, slows economic growth.

Jacob Rees-Mogg and Liz Truss have both noted this in recent days, triggering various articles accusing them of being ideologues and failing to consider the evidence.

So let’s be clear: there is very widespread evidence, compiled from around the world over the past four decades by entirely mainstream academics and research bodies (such as the European Central Bank and the Institute for Fiscal Studies), that higher public spending and taxation, relative to GDP, is associated with slower growth.

The practitioners’ rule of thumb is that every additional ten percentage points of GDP rise in public spending cuts GDP growth by around 1 to 1.5 per cent. So if the economy grew at, say, 2.5 per cent per year with public spending at 30 per cent of GDP, it might only grow at 1.5 or even 1 per cent if public spending were 40 per cent of GDP.

It is worth pondering for a moment why we should expect this to be the case. When governments tax and spend, they force people to do things they would not choose to do for themselves. This creates a number of distortions to behaviour that we should expect to result in lower growth. First, taking resources from people in taxation means they have less for themselves to spend on investing or consuming (thereby stimulating others to invest to supply for their consumption).

At very low levels of public spending, the things governments spend on, such as maintaining order, protecting contract and property rights, improving regulation and combatting monopolies, may tend to overcome coordination problems and generally facilitate economic activity in a way that means overall growth is faster.

But at some point (generally estimated to occur at somewhere between 15 and 25 per cent of GDP) government spending switches from being growth-promoting to, instead, being welfare-promoting, making the world a nicer place by helping the old and sick, sponsoring opera and sport, educating children and so on. That sort of spending, not being on what people would choose for themselves as individuals, tends to be relatively inefficient at promoting future growth and stimulating investment and innovation.

Furthermore, the process of shifting the money around is costly. A standard estimate is that there is a “deadweight cost of taxation” of somewhere between 10 and 50 per cent of the funds raised. Finally, when government does actually spend the money, it’s often very bad at it, crowding out what would be more efficient, growth-promoting private sector expenditure.

So at above something of of the order of 15-25 per cent of GDP, government spending starts harming growth, and each additional 10 percentage points added after that will tend to cut the GDP growth rate by around 1 to 1.5 per cent per year. Over time, lower growth in the economy will mean a much smaller economy that would otherwise be achievable, and that in turn will mean that in the long-term the economy actually probably spends less on healthcare, education and so on.

None of this, of course, means that the ideal is for government spending to be only 15-25 per cent of GDP. We get nice things for our government spending, and it’s worth sacrificing a bit of GDP growth for that. But it’s naive to imagine that we can get our increased assistance for the poor, the old, the sick, the artists and so on for free. It comes at a cost. And at some point the GDP sacrifice will not be worth making, even if we do not worry (as we should) about the ethical problems of excessive confiscation through taxes.

Government taxes are currently about 37 per cent of GDP. That’s the highest they’ve been since the mid 1980s — three decades. After many years of effort, we’ve got public spending down to a level at which the UK’s deficit is manageable, at around 38 per cent of GDP this year. It was a mistake to agree a large rise in NHS spending, but if we compound that error by raising spending in lots of other departments, that will mean slower GDP growth over the longer term and the undoing of all that we achieved in the years of austerity.

If we got government debt down a bit from its current level, we could probably survive as a higher public spending, higher taxing, slower-growing economy. But do we want to, and, as Liz Truss alludes to, is that the kind of economy people vote in Conservative governments to provide?

Andrew Lilico is an economist and political writer.