What do all contemporary cyclical process management theories have in common? OK, so I better define what cyclical process management theories are: all those management theories that have a lifecycle, like business process management, project management, software or systems development (and all the ones I can’t think of right now).
Oh, and the answer to the nerd riddle is: Francis Bacon, Walter Shewhart, and William Denning.
Who cares? Probably not a lot of people, but it fascinates me to divine how the modern world came to be enthralled with the hangover of 19th century ‘scientific management’ emanating from the Harvard Business School.
‘What the hell is he talking about?’ you think to yourself.
Fair enough. It’s not really a subject for a passing comment, except I thought about it one too many times in relation to other things to let it go undocumented, particularly since finding it online is not as easy as it ought to be.
Back to Francis Bacon. His most influential work, Novum Organum, was published in 1620. From its ruminations on inductive logic and ancient philosophy Bacon’s successors devised the scientific method which can be abstracted into a simple model of hypothesis, experiment, evaluation.
Skip to turn of 19th/20th century America, and Frederick Winslow Taylor is making a career out of ‘scientific management’ by controlling early mass production processes.
Taylor was a tireless self-promoter who managed to leave his imprint on the Harvard Business School’s MBA (see The Atlantic and my post on the history of business process management).
Enter statistician Walter Shewhart, enamoured with scientific management, and bent on demonstrating the assistance to manufacturing processes that could be derived from statistical analysis.
Shewart took the early model of scientific enquiry derived from Bacon to a new level by proposing it as a formula for the scientific management of manufacturing production, and not as a linear one-off affair, but as an iterative cycle, in which quantitative measurement using statistical methods would function to inspect output with the intention of improving the specifications for the next batch of outputs (see Figure 1).
One of Shewhart’s early admirers, William Deming, edited a series of lectures by Shewhart and adapted his methods in his later work on developing new methods for the US census, American wartime production, and then the 1947 Japanese census under American administration.
General Douglas MacArthur, frustrated at the utter ruination of Japanese infrastructure (there has to be a bitter irony in this), involved Deming in teaching quality control to Japanese industrialists. This led to Deming’s now ubiquitous iterative cycle of quality improvement (see Figure 1) being widely adopted and adapted into what became the Japanese school of quality management.
For reasons of abundance and lack of care, American industry forgot the lessons of process quality control that were so important during WWII, and became the world’s pre-eminent economic superpower despite enormously wasteful manufacturing processes.
In the 1970s Japanese manufacturers had become so successful, and the American economy so stagnant, that the American Congress dispatched a delegation of industrialists to Japan to learn about their methods, which the Japanese had developed from methods they had learnt from Deming (there’s an even bigger irony in this one).
Of course Americans can’t just adapt method. They needed to re-invent it to call it their own, which probably gave us the debacle of Reaganomics and our present state in which the global economy is teetering on the precarious edifice of rent-seeking, financialising assets already written off, and a renewed class warfare with pressure to maximise profit at the cost of falling wages rather than productivity or innovation.
But it also gave us the information economy and quality process controls no longer specifically designed for manufacturing processes.
Hence the all-purpose adaptation of the Deming cycle to any process-driven, cyclical management domain (see Figure 1).
The iterative, cyclical improvement process has a lot going for it, and can probably be blamed for ‘agile’ methodologies that lead to the release of products and services plainly not fit for purpose, on the promise that the next iteration will fix the bugs (hopefully before the agile start-up goes belly-up).
Today the emphasis on metrics and measurement of output, particularly by statistical methods, has waned. I suspect this is because ‘doing’ statistics is easy, but incredibly hard if the premise is to make meaningful observations about improving products and services. Most measurements never go past slashing jobs and productive capacity as a quick fix devised by mediocre minds.
And there we have it. Scientific management is alive and well today through the work of Deming in adapting the work of others, and leaving us with the all-pervasive process improvement cycle.