Is the UK really producing anything?
Photo credit: Andreas Levers
April 21, 2015 // By: James Meadway
Yesterday’s Financial Times has splashed the UK’s “productivity problem” across its front page. They’re talking about the unprecedented flatlining of productivity since 2007 – a cause for consternation as there hasn’t been an extended period without productivity growth in this country since at least the Industrial Revolution.
The FT concludes that “Lawyers, accountants and management consultants lie at the heart of the UK’s productivity problem.” A quarter of the decline in productivity since 2008 can apparently be attributed to these “professional services”. Just four sectors, “professional services, telecommunications and computing, banking and finance and manufacturing” account for the majority of the stagnation.
Productivity, defined here as the total amount produced divided by the total number of hours worked, matters. Currently, and most pressingly, it sets a barrier to wage rises. If productivity is rising, it is relatively easier for firms to pay higher wages without also eating into their own profits. But if it is flat or falling, the only way wages and salaries can increase is if there is redistribution from capital to labour.
That isn’t going to happen under this government, and, with trade unions so weak, it isn’t going to happen in the workplaces, either. So the flatlining of productivity has turned into a direct pressure on wages.
But what’s going on? I think there are two parts to the problem. The first was highlighted by the Office for National Statistics in a recent issue of their Economic Review. It shows the growth in capital stock, hours worked, and capital stock per worker over the last decade:
Capital matters to productivity. Through investment, firms (and sometimes the government) hope to replace old equipment and property, and install new. By installing new equipment, they have access to newer technology, which will generally operate more efficiently than the old. If companies don’t invest, they won’t gain access to those productivity improvements.
What the graph shows is the steady growth in capital stock (in red), alongside the ups and downs of total hours worked. The recession, over 2008-9, is very clear: a sudden drop in hours worked, followed by a recovery. The line to pay particular attention to, however, is marked in green – capital stock per hour worked. This turned negative in 2012 and has remained so until 2013, the last point for which data is available.
In other words, the amount of capital available for every hour worked is declining. While hours have picked up, companies have not invested at the same rate. And if we break down the aggregate, we find (according to ONS) that this decline was “notably apparent in real estate, professional, and information & communications services”. Manufacturing, too, has seen a decline in capital per hour worked.
The graph below, also taken from ONS, shows the relationship between changes in the capital stock, allowing for depreciation, and GVA per hour worked for the 25 largest sectors. Or, in slightly more plain English, it shows changes in the amount of capital brought to bear in each hour worked on the horizontal axis over 2010-2013. The vertical axis shows the amout of output (“GVA”) produced, on average, for every hour worked in 2013. It is showing you a relationship between net investment, relative to hours worked, and output, relative to hours worked:
The relationship is not exact. But in general, there is a close link between falling capital stock, and relatively lower productivity. This is just a simple correlation. It doesn’t, by itself, saying anything about the causes of the relationship. To find such a close relationship on a simple correlation is, however, indicative. Many industries are seeing declining productivity, counterbalanced by some seeing improvements. What ties both together seems to be the size of the capital they bring to bear for each hour worked.
This suggests a simple fix. Investment, currently falling in the UK, needs to rise. But there’s a further issue, not considered by ONS or the FT. The measure of productivity we have here has its own problems. Broadly, it is aiming to capture how much output is sold by an industry, relative to the amount of work being performed in that industry.
Because it is a measure of output sold, however, it is not quite the same as output produced. The latter would be closer to a kind of physical measure of productivity; obviously, this would not be too much use with (say) service industries like management consultants. (What, in physical terms, do management consultants produce?) Nor would we be easily able to compare the outputs of different industries: would an increase in (say) mushrooms farmed per hour worked by comparable, in any reasonable way, to an increase in car produced per hour worked?
So we have to use a measure that allows us to compare across industries, and we use value of output. This, however, contains problems. Turning again to the ONS figures, we can see the fastest growing parts of the economy: over the last year, these were construction, followed by “business services and finance”. Within that category, the fastest growing sectors were – surprise, surprise – professional services, computing, and real estate.
In other words, the activities with the worst productivity performance were, at the same time, the fastest growing. The simplest explanation for this is that firms are switching labour for capital. They are avoiding making new investments, but using more labour instead. Productivity per hour worked can decline, but total output can (seemingly) rise.
It’s real estate that is particularly intriguing. This has seemingly robust productivity growth, and rapid output growth. An industry of the future? Well, possibly.
“Real estate” is not just estate agents, and commercial property companies. It also includes rents paid to landlords and, even more strikingly, “imputed” rents from owner-occupied housing. This means ONS look at the value of owner-occupied housing, and calculate a figure as if rent was being paid on the house.
So house price rises automatically boost “productivity” in real estate. No extra activity is taking place; no new homes need be built. A simple increase in house prices is enough to generate a productivity boost. The effect is not insignificant. Stripping out real estate reduces total services productivity growth over 2014 to very close to zero.
A similar logic seems to apply elsewhere in services. The finance sector, which contributed mightily to growth in measured productivity before the crash, is another good example where “productivity” can be improved as a side-effect of activity taking place elsewhere: in this case, most notably, through the inclusion of interest payments to financial institutions as part of their measured output.
The ominous thought, lurking under this, is that strip out the more questionable productivity improvement – ones where rising asset values can produce apparent growth – and you are left with an economy that looks anything other than healthy: low investment, low productivity, low wages.
Low investment, low productivity and low wages: the UK economy is anything but healthy