Reinhart and Rogoff vs Herndon: the austerity maths doesn’t add up – but the theory does

As one of the many journalists who took Reinhart and Rogoff’s work to be a brilliant empirical underpinning for what appeared to be common-sense – that you simply cannot tax and borrow your way out of chronically unbalanced public finances – the discovery that it was, well, flawed, is something of a disappointment.

It is tempting to see this as an all-too real counterpart to that great spoof headline “Archduke Franz Ferdinand Found Alive – Great War a Mistake”: as in “economists’ backing for reducing government borrowing found wanting – global austerity programmes a mistake”. Except that, whatever went wrong with the spreadsheets, what Reinhart and Rogoff discovered was indeed right.

It is impossible to tax your way out of a chronic level of debt run up by a government. There does come a point when taxing and borrowing no longer yields the results governments seek. Taxing too much destroys incentives for everyone; pre-empts too much of a nation’s resources away from the productive parts of the economy; and the burden can indeed become so high that it becomes impossible for any tax take to service the debt. In that situation you end up borrowing just to pay the interest. High debt punishes the ability of an economy to grow and repay what it owes. The only other way out is via inflation (the usual British option). This is not and never will be sustainable.

Of course the British economy is an example of one that laboured under vast debts – 100 to 200 per cent-plus of national income – for the three decades after the Second World War. And it is true that in those years the UK did grow at a respectable pace – 2.5 per cent a year or so. But our decline was relative to other nations in those years, and by the 1970s, as all the Thatcher obsequies reminded us, it was becoming untenable. Our debt frustrated all attempts to make Britain grow as fast as her competitors. It is doing so again now. It is a truth that should be universally acknowledged, and we do not need a spreadsheet to prove it.