Why digital transformation won’t solve the productivity paradox

The productivity paradox asks the question; why doesn’t productivity increase when we invest in new technologies?

The electrification of steam-powered factories in the nineteen hundreds should have yielded huge productivity gains. Factories no longer had to be located near water or rely on shipments of coal to run a single massive steam engine, so they should have increased their outputs. But they didn’t. The factory managers of the time were familiar with how steam-powered factories operated, and so all they did was use electricity the same way they had used steam. A new technology but very little change. It was a generation later, when new managers took over and redesigned the architecture of the buildings, how the factory was laid-out, installed smaller machines, created new workflow processes, retrained workers, etc., etc., that factories realised the benefits of electricity. This wholesale change of everything that made up the factory required a new way of thinking, but still the factories failed to produce more goods more quickly. The productivity problem persisted.

In the 1970’s and 80’s, companies replaced typewriters with computers and ledgers with spreadsheets. And in the 90’s and 2000’s, the invention of the internet, cloud computing, faster processors, new software development methodologies, remote working, so on and so on, all promised to make work more efficient. The economist Robert Solow argued that companies could increase productivity purely through introducing “labour-saving machinery to reduce labour cost”, but in fact what happened was more advanced technologies required higher-skilled employees who demanded higher salaries, and so the labour cost increased but the productivity didn’t.

New Growth theory, with its emphasis on knowledge creation and entrepreneurship, argues that technology can only ever produce limited growth for a company, but that knowledge is an intellectual asset that enables increased productivity as it is non-rival and non-excludable, meaning the value extracted is not restricted by the value of the asset. That’s why software, the ultimate corporate intellectual asset, is big business. All software is fundamentally an attempt to solve the productivity paradox. Has it done it yet? Well, no.

Today we ask, will artificial intelligence and automation technologies take our jobs? And companies ask, will removing people from the process increase productivity? If you work in a certain type of role, one that requires following predictable, codifiable processes which can be automated, and your role pays enough that it’s cost-effective to automate, then yes, those types of jobs will be replaced by AI. Jobs that either require uniquely human traits such as creativity and compassion or where the labor is cheap such as delivery driving, won’t be replaced. The AI that does replace some jobs will create more jobs, and many will be even more specialist and higher paying.

So you see the paradox. But there’s another problem.

Organisations are designed to create and maintain stability. They are designed to actively resist change in favour of stock prices and employee retention. Because the structures and systems they have in place are not designed to change, they struggle to keep pace with digital transformation across sectors, industries and all of society. When technology changes quickly, organisations change slowly, and mindsets change really slowly, we get what Stewart Brand from the Long Now Foundation, calls pacing layers. If we wanted an imaginary timescale to visualise the different layers we could say technology changes in a matter of years, organisations take decades to change, and mindsets last for centuries. That’s the kind of difference in scale we’re talking about, and that’s why change doesn’t occur on a schedule. Digital transformation has to happen across all three layers, and that takes longer than we’d like.

Just as we saw with the factory managers of their day, our current mindsets and organisations struggle to adapt to emerging technologies and the new ways of thinking about technologies that are necessary for digital transformation, quickly enough. We still think of computers as a box on our desks that we interact with through our eyes and fingers. We don’t think of computers as sensor-driven AI at-the-edge monitoring river levels and controlling dams, or as a decentralized, distributed computing platform made up of lots of machines like Ethereum.

But not everyone thinks in old ways. There are those early adopters of emerging technology like AI, 5G, Blockchain, Internet of Things, 3D printing. These companies are not only leading the way, they are setting the pace of change. And it’s only getting faster.

All of which leads us to an interesting question: If there’s no productivity benefit for organisations because new technology does not lead to more output, and change is so difficult because organisations are designed to be stable, then why are so many companies trying to digitally transformation themselves? Do they even know why? Is digital transformation a fools errand of outdated organisations trying to keep up? Just as we see new technologies replacing old, is the only way for digital transformation to occur for new organisations to replace the old, and for new ways of thinking to replace outdated mindsets?

But even then, digital transformation still won’t solve the productivity paradox.