One reason for the UK's current budget deficit is the growth in welfare spending in the past 10 years, says the government. In his regular Go Figure column, Michael Blastland reveals it's rarely been under control in 50 years.
"Up 40% in 10 years!"
It's what I call a killer fact. Killer facts are numbers so obviously enormous, so scary, so… aarrghhh! that they are all we need to know.
"Up 40% in 10 years" is the UK welfare bill, in real terms, a killer fact that roared through the media. Some said it was worse: 45%, depending how it's measured and defined. Either way, a figure so enormous, so scary, so… etc.
Coalition ministers say welfare growth is "out of control". And that sentiment has become totemic to the argument about public spending. The quality press lament the figures. Everyone knows it to be true and terrible.
Erm… data, anyone?
What's meant by "welfare" is up for grabs. This chart shows the rolling 10-year total for the percentage growth of social security spending in the UK in real terms, according to figures from the IFS.
The source data used by the IFS refers to this as "net social benefits". And the figure for 2009-10 is indeed about 40% - and this figure is how much welfare spending increased over the ten years up to that date.
And so on, back to 1958-59, which shows the 10-year total from 1948-49 to 1958-59, which is also about 40%, and which takes us almost back to the beginning of the post-war welfare state.
If we take the whole period in which Labour was in office rather than just the last 10 years, the increase is about 32%, from about £145bn in 1996-7, to about £192bn in 2009-10, again in real terms.
But looking at the chart, here, for me, is the really interesting question. If a 40% - or even 45% - rise is out of control, when was welfare not out of control?
The short answer is: not often, if at all. A few years here or there, maybe. So perhaps 40%, or even 45%, given the rise in GDP, given an ageing population and the proportion of welfare that goes on pensions, given rising unemployment in a recession, isn't as insane as it first seems. But everyone knows it's gone wildly out of control, so that can't possibly be true.
And we can see that in the last couple of years, the rolling 10-year total really has gone up quickly. Something to do with the recession perhaps? So let's look at the annual increases, to see how this recession compares with previous recessions, to see if welfare in this one is more "out of control" than those in the past, in the early 1990s, early 80s and mid-70s.
This time, we're looking at the increases in each year, rather than over 10 years.
Make your own call.
All this can be said while still wondering if there is a different kind of welfare spending problem: the fact that it has risen fast and persistently for almost the whole 60-year history of the modern welfare state.
Why? That's a harder, more interesting question that we need to get to grips with. If we don't know why it happened but simply say it went out of control, we trivialise the problem and reduce our chances of doing anything coherent to address it.
It's worth saying that all governments make this kind of statement. The last one was known for rolling-up spending over a period of years and trumpeting a massive number, a killer fact, out of context, naturally.
The best antidote to killer facts? Daft questions. Like, "if it's 40% now, what was it before?" Or "is 40% a big number?" It's not hard, really.
Sources: 1948-49 to 2008-09: from ONS, Financial Statistics Freestanding, table 2.3C, then click on series ANLY. (NB these are cash totals.)
2009-10 to 2010-11: Budget 2010 Supplementary Material Table 2.8 "Net Social Benefits". (NB figures for 2010-11 are Treasury forecasts.)