Benefit spending: a quick history

Declan Gaffney has pointed out that the Daily Mail's claim that benefit spending accounts for 24.2% of "Britain's total income" is as truthful as we'd expect from the Mail - which is to say, utter bollocks. My chart shows the true picture; benefits data come from the DWP and GDP from ONS*.

This shows that the share of benefit spending in GDP rose steadily from the late 40s to early 80s, but has since levelled off. In fact, benefit spending in 2011-12 accounted for 10.4% of GDP, less than in the mid-80s or mid-90s.

Note too that benefit spending as a share of GDP fell during the last Labour government.

Also, over half of benefit spending goes to pensioners. The state pension, pension credit and winter fuel payments now account for 5.5% of GDP, compared to 4% in 1997 and less than 2% in the early 50s. This reflects less miserly pensions, plus an ageing population.

Also, since the early 80s, benefit spending has been strongly counter-cyclical. It rose in the early 80s as unemployment rose, fell during the late 80s boom, rose again in the early 90s recession, and fell during the long expansion of the 90s and 00s - a fall which has since been reversed. This suggests that what Keynes said in 1933 is still relevant today: "Look after unemployment and the budget will look after itself."

* Quarterly GDP data are only available from 1955. I've estimated financial year data pre-55 by interpolating annual data.

Comments

[Gerard O'Neill]

I can't vouch for the Daily Mail or ONS, but according to Eurostat social spending in the UK was 28% of GDP in 2010 (down from 28.9% in 2009):

http://epp.eurostat.ec.europa.eu/cache/ITY_PUBLIC/3-27112012-AP/EN/3-27112012-AP-EN.PDF

[Left Outside]

I was convinced until second reading you were responding to a particularly surprisingly clever Dean Gaffney tweet.

[Shinsei1967]

"Declan Gaffney has pointed out that the Daily Mail's claim that benefit spending accounts for 24.2% of "Britain's total income" is as truthful as we'd expect from the Mail - which is to say, utter bollocks."

Well, utter bollocks is a bit of an exaggeration.

The 24.2% is the amount of total benefits spending (incl pensions) as a percentage of government tax revenues.

Now I wouldn't use the phrase "total income" but it was fairly clear what the author meant.

[Declan Gaffney]

Gerard- Eurostat 'social spending' is not the same as benefits spending, obviously. Latter is DWP benefits + HMRC benefits, about 13% of GDP.

Shinsie 1967 24% is percentage of public expenditure, not tax revenues (cos we have a deficit, if you remember). As public expenditure is not and never has been 100% of national income, the Mail article was simply fallacious, though I think the author probably just made a mistake.

[Simon]

The psychology of how articles against deficit spending like this work is interesting to me. It seems that publications like the Daily Mail are encouraged by the rich to make the middle class hate the poor.

[Keith]

Simon is correct. What everyone needs to remember is that mailly graph type articles are lies intended to comfort the comfortable. The owners of the mail live as tax exiles in France in a Château as they are so patriotic. Which is why they hate the EU and Johnny Foreigner. Why they choose to live among the foreign types they affect to hate they do not explain. May be they like to live nears the people who vote Front National to relive the happy days when they told people Hitler was a good chap much misunderstood.

You need to look at the facts rather than the hypocritical codswallop.

[Sam]

Declan: Trying to claim that Working Tax Credit is somehow qualitatively different from Income Support because it's administered by HMRC rather than DWP is nonsense. The fact that a benefit is presented as a tax credit rather than a cheque is a completely artificial choice.

[Declan Gaffney]

Sam: what's your point? I said the figure for 'welfare' spending included DWP benefits and HMRC benefits. What is the 'claim' you're accusing me of making?

[shtove]

Tax credits included?

[Declan Gaffney]

HMRC benefits are tax credits and from 2003/4, child benefit

[Penny L]

Most benefit spending becomes someone else's income within a few days, often within the same community. Disability benefits and pensions are increasingly being claimed back as charges for council services, again providing local jobs for people who then pay tax. They are not enriching claimants in any sense. Means testing is expensive and often leads to under-claiming. Income tax should be the main way of funding services and leaving the poorest with as much as possible

[Nigel Sedgwick]

Oh dear; statistics is so difficult, and government statistics is even worse.

I have two doubts about Chris's figures, and provide links below to an alternative source with additional explanation.

My first doubt is that Chris's graph of "total welfare" does not seem to acknowledge the change is classification introduced by the UK Government in 1993. In this, after 1993 pensions (I assume the state pension) were accounted for separately; this led to a sudden drop in the government expenditure category of "Welfare". This is shown in both of my linked plots as the replacement of some of the blue bit by the red bit.

My second concern is the normalisation as a percentage of GDP. Although this is a common normalisation, and has something going for it, it confuses the issue somewhat when there are marked changes in GDP in booms and recessions. In addition, all government spending adds (one for one) to GDP; thus any change in welfare etc appears in both the numerator and denominator of the %GDP calculation. This is because GDP classifies all government expenditure as production, irrespective of how productive it actually is.

The links to the graphs follow, with my additional annotation (as the linked website unfortunately has problems with the titles for its vertical axis). In both cases, welfare expenditure is in blue and pension expenditure is in red, and they are stacked so the sum is the top of the coloured portion, both before and after the split in 1993.

Welfare/pension spending from 1950 to 2012, as % of GDP: http://www.ukpublicspending.co.uk/spending_chart_1950_2012UKp_12c1li011lcn_40t00t

Welfare/pension spending from 1950 to 2012, normalised to be per head of population and inflation indexed (to 2005 £): http://www.ukpublicspending.co.uk/spending_chart_1950_2012UKd_12c1li011lcn_40t00t

There is not a lot of reduction during the period of the last Labour government in the second plot: for either "welfare" or for the sum of that with "pensions". It is likely that at least some of the reduction in welfare payments over the post-1993 period was down to different means testing of some elderly people, after a larger than inflation increase in the state pension.

Best regards

[chris]

@ Nigel - my figures do include state pensions; they're the largest element (£74.2bn of £158.5bn of spending in 2011-12)
The figures exclude tax credits, but I don't think those who whine about the benefits bill do so because they think tax credits are too generous.
I take your point about GDP. But this cuts both ways; surely the fact that the share of benefits in GDP isn't astronomical even when GDP is depressed tells us something.

[Sam]

Declan: I apologise: lack of sleep caused me to read your Dec 1 comment exactly backward. I see now that you do explicitly make the point that some benefits are routed through HMRC and the tax system.

[Declan Gaffney]

Chris I think HMRC benefits are important part of story- some tax credit spending is substituting for earlier DWP benefits (family credit, child additions to out-of-work benefits) so we get an artefactually exaggerated fall in DWP spend in late 90's; more important though is transfer of child benefit to HMRC in 2003/4. It remains the case that spending fell sharply from mid-90's and then had longest period of stability ever 2000-2008. It's a pain that government doesn't produce an integrated series for this.