A low share of government consumption in GDP does not mean a low share of total government spending. The OBR expects the latter to fall to 38.4 per cent of GDP in 2018-19. This is higher than it was in 1997-2001. The difference between the two comprises public investment, debt interest payments and – most importantly - welfare spending.
This matters. There is reasonable - not overwhelming but reasonable - evidence that smaller government, other things equal, leads to stronger growth. But what matters here isn't government consumption but total government spending. Whilst the latter is high, two of the mechanisms whereby a small state promotes long-run growth - by ensuring low taxes and cultivating pro-market social norms - will be weak. Indeed, if you want the state to promote growth, you should prefer that it spends a lot on education and less on pensions - which is the opposite of what's planned. (Though, of course, promoting growth isn't the only role of government).
What's more, there's a positive danger in low government consumption - it adds to macroeconomic volatility. Government consumption tends to be more stable than other components of GDP - especially investment. A low share of government consumption thus makes for a more unstable economy. It's no accident that UK GDP growth was more stable after 1945 - when government was bigger - than it was before.
Of course, this wouldn't be an issue if governments could use other measures to stabilize the economy, such as fiscal or monetary policy. But it cannot do so, because downturns cannot be foreseen. Policy can therefore only respond to events. A better way to smooth economic activity is to have more government consumption: this is just what Nordic economies do.
And herein lies a reason to doubt the OBR's forecast. It envisages government consumption falling as a share of GDP because it expects investment to rise; it foresees the share of business investment (in volume terms) rising from under eight per cent of GDP now to over 10 per cent by 2019. This would take it close to the peak we saw during the tech boom.
There are two reasons to question this. One is simply that the longstanding dearth of profitable investment opportunities might not disappear so quickly. The other is that the OBR foresees only good(ish) times - of GDP growth of over two per cent. It's perfectly normal for government consumption to shrink and investment to rise as shares of GDP in such times. But we know that these only a partial sample of what really happens. When we get recession or weak growth - and it is a "when" not an "if" – government consumption will probably rise again as a share of GDP.
Whether this happens within the OBR's forecast horizon or not is a separate matter. The point is that low government consumption might not be sustainable.
