Ed Miliband will cap "structural welfare spending" if he is elected in 2015, he announced today, saying that "such a cap will alert the next Labour government to problems coming down the track and ensure that we make policy to keep the social security budget in limits."
It's not entirely clear what Miliband means by "structural" welfare spending at this point. It could just be a political fudge, designed to mimic the Conservatives' similarly fudgy focus on the "structural" deficit. The structural deficit is a particularly difficult thing to actually measure, because it relies on three pieces of information all of which are themselves uncertain: the output gap, the relation of public spending to economic growth, and the response of tax revenues to both.
Get any of them wrong, and your estimate of the structural deficit is off; get all three wrong, and you can be billions of pounds off the mark. And look at just the variation in the estimates of the output gap, via Touchstone:
But whereas the structural deficit is at least a conventional economic concept, albeit one hugely prone to measurement error, it's not entirely clear what "structural" welfare spending is, and even less clear how to cap it.
The best guess is that the structural welfare bill is the bill which we would expect to see in normal times; in other words, Labour won't view a rise in housing benefit due to the recession as a problem, but would be more concerned if, during the recovery, it fails to drop down to lower levels.
If that is the definition, then it has an interesting outcome once Labour start to cap it, because, as Declan Gaffney showed last month, "there has been no structural increase in the level of aggregate working age welfare spending for a very long time". That's because welfare spending, properly construed, must take into account foregone revenue as well as public expenditure: the most prominent example of which is tax credits. Around £3bn of the cost of tax credits in 2012/13 came from an offset to income tax. Money wasn't being "spent", but it was certainly a cost of welfare.
And when you take into account other taxation expenditures – like the mortgage interest tax relief, which was abolished in 2000 – you find that structural welfare costs have stayed remarkably stable. This chart again from Declan Gaffney's piece:
The real question is what "structural" welfare means for people not of working age. Because, thanks to our ageing nation, the state pension liability is growing year-on-year, and even pushing back the pension age by a year from 2026 won't help too much. Of course, it would be possible for Labour to define that increase as something other than structural – "demographic", perhaps – and thereby dodge the question. But if they don't, the key effect of this promise could be that Labour has pledged to cut pensions, two years before a general election against a party which has pledged to keep them above inflation and wage rises indefinitely.