25 November 2017
WITH this Budget the OBR moved from its intended status as an independent ‘watchdog’ providing uncontroversial forecasts of the public finances, into the spotlight as the purveyor of a forecast widely regarded as biased downwards in a serious way for a serious period, the next five years; so scuppering the well-meaning hopes of Philip Hammond to provide an upbeat Budget.
It has told all and sundry that this was forced on it by the behaviour of productivity growth underperformance.
However I have my doubts about this in the first place. The OBR’s macroeconomic model barely mentions productivity.
Yes, there are unit costs where it enters to help determine prices and there is an assumed ‘trend output’ where it presumably enters indirectly into the judgement about what this should be.
But there is a weak connection between trend output and the economy in their model because there is a particularly weak relation between ‘excess demand’ and inflation. This relation only operates via wages inflation.
It is weak anyway and at the present time is of course hardly operating and so may well have been overridden by the OBR forecasters; this also means that unit labour costs have been weak in contributing to inflation.
So trend output and productivity seem to have had no effect on inflation which is the only way they have an effect in the OBR model.
So how does the OBR model determine output? From demand!
In fact this model dates back to 1970 and is the same old Keynesian model that I also worked with in the Treasury a few years later!
Some may be familiar with the Item Club forecasts which use a variant of this model.
The Keynesian set-up is popular with forecasters because they can build up their view of the demand side of the economy-consumption and investment spending, exports, and government spending and then add them up to get output growth.
Supply is determined by demand. Forget expectations of inflation or the exchange rate. All such things are determined implicitly by the past.
This sort of ‘Old Keynesian’ model may be convenient but it is most unlikely to match the behaviour of the economy, not to speak of its deep violation of modern theory.
But leave that on one side: the point here is that the OBR has used its assumptions about demand to create a forecast of output, which has nothing at all to do with productivity.
These demand assumptions seem to be heavily influenced by the Treasury’s own views of Brexit: that it is damaging to consumption, investment and exports; and that government spending must be held back also to prevent more damage to the public finances. All of this is a recipe for dismal growth.
The OBR says it has been ‘neutral’ in its view of Brexit, by which it seems to mean that it has reduced export forecast growth but also reduced import forecast growth by a similar amount due to Brexit; so the two offset each other in their effect on demand for UK output.
But this is disingenuous because in these models exports are a ‘driver’ of demand from the outside while imports respond to demand inside the economy.
As a result their export assumption due to Brexit is lowering output growth and in so doing also lowering import growth: not ‘neutral’ at all!
But suppose we take the OBR’s story about productivity at face value. It is a flimsy thing to use for a forecast u-turn like this.
Productivity responds strongly to the economy as you can easily see from any ONS chart.
In recessions it plunges because output falls faster than firms lay off workers. It did so particularly badly in the financial crisis recession.
Firms hang on to a lot of workers even when output falls because workers these days are like capital: they are often hard to find again and their knowledge is vital to the firm.
Then also in recessions in a flexible wage economy like ours they become a lot cheaper to hire because they want to keep jobs and to be hired where they don’t have jobs.
In the recovery jobs rise as fast or even faster than output because of this.
But at a certain point the labour market gets tight and firms have to use their workers more intensively, just like they raise their utilisation of capital as demand rises.
For this reason even measuring ‘underlying productivity’ in manufacturing is hard because you have to allow for this variable utilisation.
Then on top of this factor is our total inability to measure quality of services, now 80% of our economy; and you realise we know next to nothing about true productivity.
The ONS is bravely trying to measure ‘hedonic’ prices of key goods like computers and mobile phones; it has an institute to measure public services quality which so far is in the discussion phase.
And nothing at all for the other 60% of GDP in the private services sector where it is likely productivity has soared .
Even on our current poor measures productivity seems to have risen sharply in the third quarter, as employment among 16-64 year olds fell 0.1% while the ONS first estimate for output growth looks like being 0.4%.
This 0.5% rise in productivity in the third quarter is an annualised 2% growth rate!
We may well be on the verge of seeing a shortage of labour developing that will force firms to use their labour stock more intensively.
It is darkest before the dawn. How ironic if the OBR has shot us, the Chancellor, and itself in the foot just when the cavalry was coming over the hill!