Sunday, April 26, 2015
English French German Spanish

Noise and Signals (April 2008)

This article was first published on The Guardian's Comment is Free

(This article was jointly written by Christopher Harvie and Noel Spare)

The credit crash can create new options - such as unlearning a lot of what has been fashionable and disastrous


In the spring of 1845 Irish cottars opened the ‘clamps’ of earth under which the previous season's potatoes were stored. They found slime. The potato blight had wiped out the monoculture, devastating their society and branding the United Kingdom, as incompetent and callous. Since late 2006 a virus as deadly has struck the financial services the UK has made its own monoculture.

In place of the innovative manufacture Gordon Brown advocated in Where there is Greed in 1989, we got ‘financial engineering’ from the City. When problems began, we were told the good money far outweighed the bad. But the first was devalued by the second, and billions (more than the value of the country's railways) have bailed out a medium-sized mortgage bank. As the panic increases, business organisations don't repair themselves but prey on one another.

This black farce was called a ‘knowledge-based economy’, but the roots of the latter were quite different in the theories Walter A Shewhart and W Edwards Deming evolved in 1920s Chicago about what understanding systems and managing processes really meant. Shewhart's ideas became a useful link between economic and political decision-making, notably and effectively in postwar Japan.

Shewhart was a product of American ‘pragmatism’: the philosophy of C S Peirce, William James, John Dewey and CI Lewis. Derived partly from the Scottish common sense school of the 18th century, this was different from the rigidity of F W Taylor's ‘scientific management’, stressing instead practical systems analysis and problem-solving. In the 1920s, faced with an inefficient telephone works, Shewhart stressed that in order to understand system behaviour one had first to analyse its measurable characteristics (data) to determine its capability. This involved distinguishing between ‘noise’ (routine variations in the system) and ‘signals’.

‘Signals’ were outside routine systemic behaviour but had causes which were ‘assignable’ and treatable by a linear (cause and effect) approach. Shewhart discovered the virus of variation but most importantly, showed that it existed in two strains, and each required a different treatment. Confusion of one with the other would lead to inappropriate treatment and disastrous consequences.

Noise in a system could not be cured with a linear approach, ie by assigning causes to discrete measurements. This was ‘tampering’ and made things worse. In fact, Shewhart's most profound finding was that once signals are eliminated, no further systemic improvements are possible. Curing a noisy system meant changing it in a fundamental - and co-operative - way.

Applying this approach to contemporary Britain, noise is reaching the pain threshold with infrastructure crises, poor productivity and social misbehaviour endemic. Worrying enough, with a higher education system churning out ‘rental’ jobs in the law, public service and big firm bureaucracies, and not technical competence which is always under proof. The current level of noise in Britain-as-a-system ought to demand fundamental re-engineering. But can this diagnosis be made without adequate educated and trained manpower? Without it, will outcome-obsessed politicians misinterpret noise as signals and - to show initiative and capture publicity - try to ‘get more out of the system’ by pressurising people, usually by imposing what Shewhart/Deming called ‘arbitrary numerical targets’. When these are, in an unaltered system, impossible to meet, will the result turn malfunctions into complete collapse?

This has been evident in the turbulent history of Blair-Brown bureaucracy. Think of (1) the displacement in our schools of humane education by testing, resulting in intellectual overstress and imaginative aridity. Or of (2) the unending succession of computer crises - mega-schemes that don't work - in the public service. Or of (3) catastrophes in the non-computer area: in railway upgrades or military equipment for Iraq or London air terminals. Or of (4) scandalously target-driven healthcare. We were once supposed to be good at software and incorporating feedback: if we fail there, can the rest be any better?

Systems theory: pragmatism: common sense. Gordon Brown got this ancestry right. But appreciating Adam Smith was his limit. Cool assessment and a distrust of self-interested rhetoric escaped him. Had we, in a supposedly steerable mixed economy, devolved information to the people with the technical competence and responsibility to steer? When the trade unions, a type of self-management, with an interest in training (still the case in Germany) had been made sidelined in favour of consultants and executives?

And were they too close to Smith's ‘conspiracies of merchants’ with a vested interest in dysfunction? During the North Sea oil boom in the 1970s, the Seven Sisters simply outbid their supposed regulators (state salaries could never match the private sector). Financiers close to government cleaned up in the ‘wild capitalism’ phase of Thatcherite privatisation, sold out and cleared off. Look where Brown's ‘light-touch regulation’ in the City has landed us.

Brown/Blair emphasised market and management, but were the two compatible? What if the ‘churn and burn’ of turbocapitalism destroyed the work-community within which intelligent management had to operate? What if management as an ‘interest’ not a ‘brain’ used this chaos to overstate its competence: to increase demand, and hence payment? Is there a point where the run-down of ‘traditional metal-bashing’ carries industrial rationality - and the ‘high-value-added services’ - away with it?

Now, shocks have dislodged what stability remains. But credit crash and political disruption can create new patterns and options. Our ways out - for example through renewable technology - will involve re-learning how firms and governments cohere and renew themselves; and unlearning a lot of what has been fashionable and disastrous.

Home > Comment > Noise and Signals (April 2008)