There is concern among some organisations that British manufacturing is seen as a poor relation to the service sector because it contributes less to national output. ‘Manufacturing is talked about in a derogatory way, because the service sector is perceived as three times as big,’ says Liz Amos, director of the Institute for Manufacturing.
The Institute argues that government figures do not take sufficient account of the growing integration between manufacturing and services. It believes the complexity of company operations for example the increasing popularity of outsourcing makes value added by various activities harder to measure.
A manufacturing site may support many peripheral service-sector jobs as well as those directly created by the factory, so statistics understate the significance of the manufacturing sector to the economy as a whole. The interrelationships are not well understood.
The institute sees the treatment of some sectors in official figures as anomalous: information technology is counted as a service, for example.
Also, many people believe the official statistics, which aggregate figures for the sector as a whole, give less insight than data at the level of an individual plant but these are not published.
‘There is a strong school of thought that you need to start at company and plant level to help inform the debate,’ Amos says.
She says the problem is not just about gathering statistics. ‘It is a question of understanding what is happening and we are not very good at that.’
There are significant implications for government policy. This was emphasised when research by the London Business School (LBS) questioned whether the Bank of England was basing interest rate decisions on misleading official statistics which underestimated productivity in manufacturing.
Researchers Neil Blake and Paul Robson used CBI survey data, and came up with stronger growth for manufacturing than the official figures. For 1997, the LBS research showed productivity rising by 3.2%, against the Government’s claimed 1.2% which meant the fears of the Bank’s monetary policy committee that earnings growth was in danger of fuelling inflation were unfounded.
Blake and Robson argue that the Office of National Statistics used employment figures which distorted the level of manufacturing productivity. But they say things have improved. ‘They are now using better technology to collect the data,’ Robson says.
The LBS team admits the manufacturing sector is hard to define, and what it produces is hard to quantify. The true output figure may be somewhere between the LBS and ONS levels, Robson suggests.
The Bank of England and DTI have some sympathy for the business-school team’s viewpoint. ONS has also had ‘a hard look at the issues raised’. according to James Partington, its head of employment statistics from employer surveys. But officials cannot fully explain the discrepancy.
According to Alan Armitage, chief economist at the Engineering Employers’ Federation, the complexities of defining manufacturing and measuring what it produces are immense. He grappled with the problem two decades ago at the National Economic Development Council. ‘We came to the conclusion that looking at individual sectors plant by plant was the only way to start removing distortions,’ he recalls.
This is still a favoured approach among many organisations. But ‘there is a huge risk of comparing apples with pears’, Armitage warns. It is also difficult to tell whether what you are looking at is typical.
There is another problem. Official statistics are standardised to meet European and international requirements. These dictate how the figures are collected, limiting room for manoeuvre.
The ONS has just overhauled its method of gathering output data, as part of a regular five-year reweighting exercise, and to introduce European standards which have to be in operation by next April. So whatever the criticism of ONS figures, its officials stress the office has been sticking to standard European practice.
Rob Knight, senior policy adviser at the CBI, agrees on the need for improvement. ‘The LBS report was timely,’ he says. ‘It showed the need for improved data, especially on the employment side.’ He adds that the CBI also believes the contribution of manufacturing to the economy has been understated.
The CBI’s National Manufacturing Council has been looking at how to assess the real contribution of manufacturing to the economy, taking account of its interrelationship with other sectors. ‘We are doing new work with the Institute of Management on the role of IT, but it is early days,’ Knight says. He says that for industry classifications to reflect the right things, ‘we have to understand interrelationships’.
ONS is working on this too, but warns: ‘The figures are not easy to unravel.’
The IT industry does not see its classification with services as a burning issue, according to John Higgins, director general of the Computer Services Association. The official definition of the sector is adequate, he says.
‘The issue is whether companies are changing the way they operate. Often manufacturers provide a product wrapped up in a total service. This has been the case in the ITsector for some time.’
Knight says another key task is to ‘make sense of the productivity debate’. To get a more accurate picture, ‘we will probably do more anecdotal and survey-based work’.
Partington says the ONS is trying to improve its measurement of productivity by moving towards a measurement in which total output is divided by the number of hours worked. At present, output is divided by the number of people employed.
ONS is also looking at wider definitions of productivity, including a measure based on dividing output by total employment costs. This is to solve the problem that if a company sheds senior staff and takes on more junior people, it makes a saving on the pay bill.
‘Companies will regard this as an efficiency gain, but it will not register in the official measure of productivity, which is currently output per head,’ says Partington.