To keep my comment section happy, here is some “productivity scepticism”. I like this theory from Paul Donovan of UBS discussed in a Guardian article, here quoting Donovan:
“In 2000, 32% of UK businesses were employers. By 2014, 24% of UK businesses were employers. This raises the obvious question of what on Earth 76% of UK businesses were doing if they were not employing anyone – and the answer of course is that they were single person businesses where the owner was the sole person ‘on payroll’.
This matters for capital spending, because people setting up as self-employed – for example, as a consultant – may well make little or no upfront investment in kit. Says Donovan:
“If you are a self-employed consultant, you probably already have a laptop, have a car, have an office at home. As the boundaries between home and work blur, we are making better use of the capital we have got.”
“What this means is that investors looking for a “capex recovery” may be missing the point. The secret capex story may be that businesses make better use of non-business assets, and that part of the capex cycle … masquerades under a ‘retail sales’ pseudonym.”
One of the puzzling parts of the UK productivity puzzle is the divergence of employment and output during late 2011 through 2012, when output barely grew and employment soared. If output and employment are both growing we can use the “low-skilled labour supply shock plus composition effect” theory to explain slow average productivity growth, but that doesn’t work well for this period.
We can (partially) resolve this problem by revising up 2011-2 real GDP, and Donovan’s theory could help here. 2011-2 saw a self-employment boom; if capital investment by the self-employed shows up only as consumption, or not at all, then measured GDP was too low. The recovery in household consumption of durable goods has been much stronger than total consumption, up 19% from 2008 Q1 to 2014 Q1 versus 2% for the latter. One of the upcoming ONS methodology changes is to remove a £500 lower bound on what counts as capital spending in the business investment survey, perhaps this will help.
File under “grasping at straws” if you prefer. There are good reasons to be sceptical there is any significant bias in measuring GDP, and other survey indicators (PMIs etc) suggest the 2011-2 quasi-depression was roughly that. Stories about the unemployed being pushed into “false” self-employment could make us doubt the jobs data instead, though again it would be surprising if that was sufficient to explain such a large shift in the labour market surveys.
Back in 2013 I described a (crude) supply-side counterfactual for UK real GDP using the OBR’s 2011 Forecast as the baseline. Here is an update which uses the data from the OBR’s June 2010 Budget forecast and extends to 2014. (I am not sure why I did not find the forecast for hours worked in the 2010 forecast before.)
The exercise is again simple (or if you want, much too simple): I take the OBR’s June 2010 forecast of output per hour, the observed data for total hours worked, and calculate a supply-side counterfactual path for real GDP. That path is compared against both the OBR’s June 2010 forecast of real GDP, and the actual data we have for for real GDP – I’ve rebased this time to highlight the difference since the pre-recession peak.
It remains true that a purely demand-side view of the UK recovery seems to prove too much. At least, it is hard to see that the weakness of RGDP relative to expectations circa June 2010 can be explained by a demand-side view; the shortfall in output is simply not a reflection of a shortfall in employment.
IFS chief Paul Johnson laments the absence of political debate about productivity:
What happens next to productivity will define our economic performance over the next five years. In the end growing productivity is the key to our economic wellbeing. As the Nobel Laureate Paul Krugman once put it, productivity isn’t everything, but in the long run it is almost everything. Strong productivity growth will lead to higher earnings, higher living standards, and an easier job reducing the deficit. If productivity growth is weak, then we are in for some more tough years.
So why don’t we hear more about this from the politicians? Whatever they may say, they really can’t just legislate for higher earnings and lower prices. Those will come only as a result of a more productive and efficient economy.
It’s funny really. Some will attribute the stagnation of living standards since 2010 (or 2008) to mostly demand-side causes (e.g. Coalition austerity), and get extremely cross that politicians don’t take the AD policy seriously… and others get extremely cross that politicians don’t take supply-side policy seriously enough. The spectrum of views between the two extremes is also available.
I get more frustrated that we lack a common methodology or understanding of how to interpret the macro data. How do we decide the extent to which each view is true? Without that, I’m not sure why we should expect politicians and policymakers do much better than pick some position. (To be fair, many fail even that test.) Crazy models like the “paradox of toil” – which are taken seriously by some academics – make a mockery of the idea we can or even should separate the supply-side from the demand-side.
Unemployment towards the end of 2014 was 0.3% higher than in early 2007. Is it unreasonable based on figures like that, to say sure, the stagnation in UK living standards has been “clearly” supply-side? I could pick other data: low inflation or wage growth maybe, low (forecast) nominal GDP growth, and say, yes, demand-side problems remain.
But it’s complicated. If productivity growth has fallen to around 0% (god forbid) then nominal wage growth at around 2% is consistent with something like “full employment”. The argument goes back and forth… it only appears satisfying if you have strongly held prior beliefs and find a data point to confirm them.
Good retail figures today from the ONS, with another decade high:
• Year-on-year estimates of the quantity bought in the retail industry continued to show growth for the 20th consecutive month. In November 2014, the quantity bought increased by 6.4% compared with November 2013. This was the highest year-on-year increase since May 2004 when it grew by 6.9%.
This is always my favourite series:
There have been quite a number of papers over recent years (e.g. Cheshire et al, 2012) documenting how UK land use regulation has damaged retail productivity, for instance by restricting store size. I wonder to what extent the shift to on-line shopping is both a response to – and ultimately a route around – those regulatory restrictions. On a per capita basis, Amazon appears to have more warehouses in the UK than any other country; it would be fascinating to know how their productivity compares across countries.
The 2008 Pre-Budget Report was a response to a combined demand-side and supply-side shock. The major discretionary changes were a temporary deficit-funded VAT cut (demand-side or supply-side stimulus, take your pick), bringing forward some planned capital spending (ditto), “efficiency saving”-type cuts planned for 2010 onwards, plus:
1. introducing a new 45% marginal tax band for high earners,
2. introducing two new 60% marginal tax bands for high earners (“for no obvious economic rationale” – Robert Chote),
3. raising employee and employer-side national insurance rates from 2011,
4. raising alcohol, tobacco and fuel duty.
Pop quiz. Would you expect those policy changes to (a) increase or (b) reduce potential aggregate supply? If you don’t like the framing of this obvious right-wing trap, let’s ask another question. In that PBR document the Treasury described the situation as follows (p161):
[…] the global credit shock has significantly increased the uncertainties surrounding trend productivity and it is therefore difficult to assess to what extent lower rates of productivity growth over the course of 2007 and 2008 are likely to reflect cyclical or structural developments. Challenges in decomposing recent developments into cyclical and structural elements are likely to persist for some time.
From the perspective of 2014, the last sentence in that quote is certainly astute; the challenge of decomposing cyclical and structural issues has arguably only got harder since 2008. My (slightly) less loaded question, then: would the aforementioned tax hikes reduce or increase your uncertainty around the future path of productivity?
I had a depressing conversation with a friend recently, a high-skilled high-ish-earner with two growing-up children and a partner looking to return to part-time work after a few years providing full-time childcare. The family had calculated that due to the insane schedule of child benefit withdrawal, they could raise their net household income by having the high-skilled parent take a day off per week, and have the lower-skilled parent take a part-time job working that day. (This happens because the increasingly generous personal allowance applies a 0% marginal rate to the lower hourly wage and the benefit withdrawal schedule imposes a 60+% reduction on the higher gross hourly wage.)
That’s only a dumb anecdote, and I hate policy-by-anecdote debates, but it seemed like a good reflection of what has happened to the UK labour market. Since 2007 the government has introduced policy after policy aimed at reducing labour supply from high-skilled workers, just a sample here:
- The top tax band, now back at 45%, remains higher than in 2007.
- Marginal tax rates due to various withdrawals have risen at 60% and above for some.
- “Fiscal drag” means an increasing proportion of the workforce is hit by higher marginal rates; the IEA claim that in 2013, one in six income tax payers face a 40% tax rate, up from one in sixteen in 1990.
- Immigration reform aimed at reducing the number of (future) high-skilled workers entering the country.
For many of the policies above, I would think we would see the “long run” effects pretty quickly. At the same time, we have benefit reform and a large increase in the personal allowance aimed at increasing labour supply from low-skilled workers. If you discourage labour supply from high-skilled workers at the same time as encouraging supply from low-skilled workers, what’s the expected effect on average productivity? I don’t want to overstate the case, but surely at least some of the reduction in productivity which we’ve seen is just not that hard to explain. (Post title stolen shamelessly from Sumner/Lucas.)
PS. Since I know some progressives believe raising tax rates causes people to work harder, I asked the Guardian’s personal finance writers for their advice to those facing higher marginal tax rates:
Taking an extra week’s unpaid holiday,
The first six words were enough.
This data is slightly better than was expected, output/hour rose in absolute terms.
It looks like 2013 Q4 is going to have very bad productivity numbers.
This post is a little premature, we need to wait for another month of labour market data, and the month 2 estimate of GDP. But the data seems likely to show that the UK has had around six quarters of fairly steady, fairly strong demand growth, averaging at least 4% NGDP growth from 2012 Q3 to 2013 Q4. Yet the level of market sector output per hour is likely to show almost no change at all over that period. This is starting to look like a total disaster for those of us hoping for at least some validation of the “endogenous aggregate supply” theory.
Chris Giles comments, as does Duncan Weldon. Duncan picks up a point which friend’o’the’blog James made in the comments here in a previous post: that average productivity is looking bad as low-productivity, low-waged workers regain employment, even if productivity might be rising for the rest of the workforce. Karl Smith makes a similar point at FT Alphaville. That’s all very reasonable, but this idea is giving in to supply-side pessimism. Supply-side optimists need to argue we have high-skilled, highly productive workforce part of which is merely sitting idle waiting for demand to return and put them back into work. But that does not appear to be happening.
It would be very interesting to see an update of the IFS analysis which showed where the loss of productivity has occurred in terms of movement of output/hour by sector and in the composition of hours between sectors.
I spent five minutes with the Labour Productivity data for 2013 Q3 and I get a feeling that I won’t like the answers. Here is a (perhaps) startling statistic: in the finance sector total hours worked is already back around the same level as 2008. Output per hour in that sector is down 12%. That hurts so, so much. Karl says, and he’s made the point before:
However, we should not forget that British productivity was supported by two very high margin wells which have begun to run dry, North Sea Oil and City of London Financial Services.
Replacing those sources of wealth will not be easy – but I add, it very likely will be done. Oil and Finance promoted a high priced pound and as a result made life difficult for other British exporters and life sweet for British importers. As that phenomenon turns around we should see a Britain that is increasingly self-sufficient and investing in technology and education to support the needs of its population rather than serving foreign demand for energy and complex derivatives.
It is an uncomfortable message to receive. To end on a positive note, I heartily endorse Sam Bowman’s “Alternative Agenda for Hope“.