Friday, January 05, 2007
Nurses More Elastic Than Doctors
As previously confessed, Tyler's first proper job back in the Soviet era was economist at the then Department of Education and Science. And one of the first projects he ever got involved in was manpower planning for teachers.
The idea was to make sure that we trained the Goldilocks number of teachers. Not too many so that we wasted loads of money training people for the dole queue. Not too few so we had to have school classes of 113. Goldilocks was just right.
How hard could that be, you ask. Presumably all we needed to do was get a forecast of the school age population from those nice if somewhat grey people down at the Government Actuary's Department, make some assumptions about pupil-teacher ratios in state schools, and Bob's your uncle.
Ah. Well. In theory, yes.
But the practice was a tad trickier. For a start, the Government Actuary turned out to be not much good at forecasting the future birth rate- even a couple of years ahead. And then, it took three or four years to train a teacher, and that could only be done in training establishments that had to be staffed up two or three years ahead of that (although admittedly, these days the whole thing probably takes no more than two weeks in one of those burger universities).
And then some uncertain proportion of those who entered training would either jack it in, or graduate but then not actually go into teaching (Tyler being a case in point). And then there was the issue of people leaving teaching after being horrifically scarred by 3c- how many should you allow for? And then there were the dreaded "married women returners"- why train new teachers when you could attract back more experienced female teachers who had taken a break to have kids? But how many?
And wouldn't that depend on relative pay?
Ah yes, relative pay.
What eventually dawned on us was that this wasn't simply a technical exercise of crunching Stalinist manpower statistics, but more what we economists rather pompously like to call a behavioural issue. Without making some assumptions about future teacher pay, we wouldn't have a prayer of working out how the players might behave, and how many teachers to train.
Of course, it was one thing to realise that, quite another to do anything about it. And the whole teacher manpower planning thing continued to stumble along in the usual haze of big reports and inconclusive meetings. For all I know, it still is.
Which brings us to that leaked Department of Health document on NHS manpower. We old skool Soviet manpower planners can see all the classic signs of terminal muddle: forecasts overtaken by events, financial stop-go, flip-flop recruitment, and wasted training. The headlines have focused on the expected 3200 consultant surplus and 14000 nurse shortage by 2011. And Reform have quite rightly been pointing out that the whole tenor of the document is that of unreformed top-down Soviet planning, rather than the supposed new world of patient choice and local decision-making.
But for us ex-planners, just as interesting is the small print (warning- next few paragraphs may not interest non-ex-planners). Because in the leaked annex (here) there's a discussion of pay elasticities:
"A wage elasticity of supply gives the percentage supply response to a change in the wage offered. Assuming a wage elasticity of 1, a 10% increase in wages will lead to a 10% increase in hours worked. For both doctors and nurses estimated elasticities are low –i.e., below 1."
What does that mean in English? It means that as you increase the relative pay of doctors and nurses, you don't get a proportionate increase in the amount of work you get out of them. So you end up paying more "per unit of work". As the employer, you get worse value.
In the case of doctors, the paper suggests the effect is particularly pronounced. Doctors seem to be highly "inelastic"- you can pay them more but you'll get virtually nothing back in terms of extra work. Which of course is pretty well what happened with those new contracts.
Nurses are a somewhat better bet. Not only is their "hours" elasticity higher (0.5), but those married women returners are also more likely to come back ("participation elasticity" of 0.55).
What does this mean for the cash-strapped NHS? Bluntly, it means that doctors and nurses- especially doctors- can expect a pretty rigorous pay squeeze from here on in. The NHS needs the money, and if the planners have convinced themselves that higher pay means worse value, the converse clearly hold true.
A 2.3% pa overall NHS pay increase is being pencilled in through to 2011: don't be surprised if it's lower, especially for doctors.
If you were a doctor on a final salary pension scheme heading towards 60, you might want to run your spreadsheet over the timing of your retirement.
PS The DoH document also proposes moving away from national pay bargaining in order to "save money". We've blogged before about the problems caused by national scales, especially in London and the South East where nurses seem to be substantially underpaid (see here and here), so the proposal does sound like a step in the right direction... even if it would mainly redistribute cash from staff up North to staff down South, rather than saving money. But will it happen? In the teeth of implaccable union hostility? Talk sense will you.
Posted by Mike D at 11:16 am