I’m suprised by this claim from an ONS economist:
The ONS estimates that, in the year to Sept-Nov, total hours worked in the economy fell by 2.5%; table 7 of this pdf. Yes, this is a full percentage point more than the drop in employment, thanks to the shift from full- to part-time working. But it’s still short of the 3.2% drop in GDP in the year to Q4.
In other words, productivity has fallen. In this sense, this has been an unusual recession. My table shows why. It shows that total hours worked actually fell less in this recession than in the 1980-81 or 1990-91 recession. But GDP fell more. This means that productivity - output per worker-hour - fell more in this recession than in the previous three.
This is odd. The labour market is supposed to be more “flexible” now than then, so firms should have been more able to cut jobs. And firms have been desperate to raise cash, so they should have needed to shed labour.
So what’s happening? Five possibilities can be dismissed or downgraded.
1. Statisticians are over-stating the fall in GDP. Maybe. But surely not that much.
2. GDP has fallen because oil output has fallen. True. But this doesn’t much narrow the gap; excluding oil, the peak-trough drop in GDP has been 5.4%.
3. Workers have priced themselves into jobs, because real wages have fallen. However, they also fell in the 1980-81 recession, but this didn‘t stop employment dropping sharply. And the price-elasticity of demand for labour can‘t be that high.
4. There’s been a shift to public sector employment. However, insofar as productivity in the public sector is measured by the wage bill, this won’t affect measured aggregate productivity; if someone loses a £20,000 job in the private sector but gets a £20,000 one in the public sector, measured productivity will barely change.
5. The high productivity (well, high-wage) finance sector has been hit hard in this recession. But this doesn’t show in the figures. The peak-trough drop in output of business services and finance has been 6.1% - the same as in the general economy.
This leaves two other non-exclusive possibilities. One is that there’s been more than usual labour-hoarding in this recession. Employment these days consists of more high-skilled workers than in the past. But employers are always loath to sack skilled staff, for fear of not being ably to rehire them in an upturn.
The other possibility, though, is that the real business cycle mob might have a point. Perhaps this recession isn’t (just) about demand shocks, but about an adverse productivity shock too. This would be consistent with the fact that inflation has not fallen as expected.
If it’s the former, then we have a good reason to fear that the upturn in GDP won’t create many jobs, but will rather just cause increased labour utilization. If it’s the latter, however, then perhaps the entire basis of macroeconomic policy has to be called into question - because, as I say elsewhere, an economy buffeted by supply shocks is one in which inflation targeting would add to the volatility of GDP.
Economists have been puzzled by the fact that GDP appeared to continue falling, or remain flat, when employment rates – which are much slower to reflect upturns in the economy than GDP – appear to have stabilised.Now, it’s certainly true that part-time employment is rising. But this doesn’t explain much of the divergence between GDP and employment.
Aileen Simkins, economist at ONS, noted that one reason may be the divergence of full and part-time employment.
Part-time employment is rising while full-time employment is falling.
The ONS estimates that, in the year to Sept-Nov, total hours worked in the economy fell by 2.5%; table 7 of this pdf. Yes, this is a full percentage point more than the drop in employment, thanks to the shift from full- to part-time working. But it’s still short of the 3.2% drop in GDP in the year to Q4.
In other words, productivity has fallen. In this sense, this has been an unusual recession. My table shows why. It shows that total hours worked actually fell less in this recession than in the 1980-81 or 1990-91 recession. But GDP fell more. This means that productivity - output per worker-hour - fell more in this recession than in the previous three.
This is odd. The labour market is supposed to be more “flexible” now than then, so firms should have been more able to cut jobs. And firms have been desperate to raise cash, so they should have needed to shed labour.
So what’s happening? Five possibilities can be dismissed or downgraded.
1. Statisticians are over-stating the fall in GDP. Maybe. But surely not that much.
2. GDP has fallen because oil output has fallen. True. But this doesn’t much narrow the gap; excluding oil, the peak-trough drop in GDP has been 5.4%.
3. Workers have priced themselves into jobs, because real wages have fallen. However, they also fell in the 1980-81 recession, but this didn‘t stop employment dropping sharply. And the price-elasticity of demand for labour can‘t be that high.
4. There’s been a shift to public sector employment. However, insofar as productivity in the public sector is measured by the wage bill, this won’t affect measured aggregate productivity; if someone loses a £20,000 job in the private sector but gets a £20,000 one in the public sector, measured productivity will barely change.
5. The high productivity (well, high-wage) finance sector has been hit hard in this recession. But this doesn’t show in the figures. The peak-trough drop in output of business services and finance has been 6.1% - the same as in the general economy.
This leaves two other non-exclusive possibilities. One is that there’s been more than usual labour-hoarding in this recession. Employment these days consists of more high-skilled workers than in the past. But employers are always loath to sack skilled staff, for fear of not being ably to rehire them in an upturn.
The other possibility, though, is that the real business cycle mob might have a point. Perhaps this recession isn’t (just) about demand shocks, but about an adverse productivity shock too. This would be consistent with the fact that inflation has not fallen as expected.
If it’s the former, then we have a good reason to fear that the upturn in GDP won’t create many jobs, but will rather just cause increased labour utilization. If it’s the latter, however, then perhaps the entire basis of macroeconomic policy has to be called into question - because, as I say elsewhere, an economy buffeted by supply shocks is one in which inflation targeting would add to the volatility of GDP.
I suspect it is a combination of the two. We've seen huge negative demand and supply shocks (in contrast to the last decade when the supply shocks were positive). This is going to be a huge headache for policymakers because it implies the growth-inflation mix has changed for the worse. If the inflation target is left as it is, then we will have to accept weaker real growth. That will be a genuine test for whoever is in Number 11 after 6 May.
Posted by: Econoclast | February 02, 2010 at 02:51 PM
Perhaps the measure of productivity is wrong. Chris acknowledges the concept of labour-hoarding, but is there productivity hoarding too? If you are confident that there will be a recovery shortly, why not deploy your workers to create a better company for the next "normal times"?
Posted by: charlieman | February 02, 2010 at 09:50 PM
One big difference between now and the early 80s is that the unions were far more powerful then. In their usual considerate way they were happier to see people laid-off rather than taking a pay cut which could bugger-up their precious industry-wide rates of pay. Now I suspect pay cuts are far more widespread in those businesses that cannot to retain all their workforce on the same wages; thereby allowing the workers to remain employed and the employer to have his workforce in place for any upturn.
Posted by: FatBigot | February 03, 2010 at 01:00 AM
Macro policy being called into question?
Who would've thunk it.
Posted by: chris c | February 03, 2010 at 08:28 AM
@Charlieman
The problem is that if the economy has misallocated resources (too many estate agents and bankers), any labour hoarding here is simply deferring the adjustment (the government subsidy to banks has encouraged this).
@FatBigot
For the economy as a whole, there is no evidence of pay cuts - though pay growth has moderated. Unfortunately, given the collapse in productivity, this has still forced up unit wage costs. That is eroding corporate profit margins and also holding up inflation.
The level of aggregate employment is too high given the economy's diminished output and potential to grow. When the banks and the public sector start to shrink, then unemployment will rise again.
Posted by: Econoclast | February 03, 2010 at 08:47 AM
Then you have to explain what the supply shock is! You are not only claiming that there is a fall in the rate of improvement of productivity, but there is a fall in the absolute productivity.
But lets just imagine that there was persistant mispricing of important classes of goods. And these goods were cheap to produce. Then observed productivity could fall dramatically, without any change in real productivity. Seems more likely to me.
Posted by: reason | February 03, 2010 at 01:48 PM
One explanation would be that the housing-financial boom resulted in a structural misallocation of resources into finance, real-estate, import-intensive consumption, and construction. While the boom lasted, this would have looked like a gain in productivity, precisely because the outputs were mispriced upwards.
We're now seeing the painful reversal of this phenomenon - redistributing resources into exporting and import-competing manufactures and services, shrinking the financial and real estate sector, and possibly underpricing its outputs (if they could be overpriced before they can be underpriced now).
Posted by: Alex | February 04, 2010 at 10:01 AM
P.S. Hint - I think the mispriced goods are financial assets and the income generated from their turnover (agency fees) is easy to produce. It also conveniently is not part of the CPI.
Posted by: reason | February 04, 2010 at 03:38 PM