Thursday, 26 February 2009

Think of a number, any number

If the number you thought of was 8,664, the chances are your name is Alex Salmond.

That's the figure the Chancer in Chief has come up with for the number of jobs lost should Scotland have to accept a budget cut of £1 billion over two years thanks to the 'efficiency savings' proposed by Alistair Darling.

If you do the calculation, that gives a figure of £115,420.13 per job. The mean average full-time earnings last year in Scotland were £28,296, while median average earnings were approximately £23,925 (weekly figure of £460.10 x 52). If we assume that all those who lose their jobs would be on average wages, then Salmond is basically assuming a multiplier effect of 4 or 4.5 times the wages lost.* That seems rather high to me.

Salmond seems to be assuming that if posts are lost, none of those involved will find alternative employment or would retire. There also seems to be an assumption that those affected would see their spending drop to something close to zero, which is very unlikely to happen. Even if only a quarter of those who lost their jobs found other work and the rest continued to spend at 25% of their previous level (on esssentials such as food and electricity), then that would lift the multiplier effect that Salmond's assuming to well over 5 times the wages lost, which seems even more unrealistic.

But of course, the Scottish Government could only have come up with such a precise figure like 8,664 if they'd done a thorough study of the services to be affected and knew exactly where the axe would fall in every department. But if they'd done such a thorough study, I'm sure they could have come up with areas where things could be done more efficiently - maybe about £1 billion over 2 years, perhaps. With a public sector as large as Scotland's, that shouldn't be that difficult.

Now, I wouldn't want you to think that Salmond had just plucked a large figure for job losses out of the air and was using it to scare people and score political points. Perish the thought. But I would be really interested to see whether anyone can justify such a precise figure as 8,664 jobs lost. I suspect the reality would be rather different - maybe more, maybe fewer jobs going. But I doubt very much it would be 8,664.

* There is of course the factor that people who lose their jobs would then be on benefits and the cost of those may be included in this figure. But equally, people on benefits would still have some spending power, so those factors may cancel each other out.


Stephen B said...

Coupla things...

The multiplier figure looks credible enough to me - it's probably taking into account direct, indirect and induced employment effects which do tend to raise these things higher. More here for those who find it difficult to sleep.

But it is scaremongering for the reasons you mention. Put simply it ignores displacement effects or a counterfactual scenario.

But any politician worth their salt is surely going to argue that more, not less, money should be spent in their area so I don't see that wht he is doing is particularly out of the ordinary.

Stephen Glenn said...

Bernard you brought a smile to this statistician's face with this.

Bernard Salmon said...

Stephen, I had a look at the link you gave me and that gives the Type 1 and Type 2 GVA (I presume gross value added) multipliers for public administration as 1.5 and 1.8 respectively. For things like education and health the numbers are even lower. I genuinely don't see where the 4-4.5 times figure comes from. Can you explain that to me (in a way a non-statistician would understand)?
I agree that Salmond's stance is particularly unusual, but that doesn't mean he shouldn't be pulled up for it and asked to explain how he came up with his figures.

Bernard Salmon said...

Sorry, I should have said Stephen B.

Bernard Salmon said...

Aaaagh, I meant Salmond's stance is NOT particularly unusual.

Liberal Democrat Blogs