Why Is Technology Not Producing Productivity Improvements?

In 1987 Robert Solow famously remarked that computing is everywhere except in the productivity statistics. His comment reflected frustration that despite the considerable hype around the digital technologies of the day, productivity stats were barely moving. It was widely believed that this was largely a case of organizations attempting to introduce new technologies to processes that were designed very much with the old ways of working in mind. It was only after business process reengineering was introduced to design organizations around computers that things began to shift.

The recent Global Innovation Index raised concerns that we’re leaving through a Solow redux, with considerable investments in technology, innovation, and entrepreneurship failing to deliver the kind of productivity improvements that improve the lot of people across society. The researchers openly pondered whether the stagnation and low productivity levels seen since the financial crisis are likely to endure or whether technologies like AI and cloud will eventually drive society forwards again.

The University of Toronto’s Avi Goldfarb, Joshua Gans, and Ajay Agrawal make the case for business process reengineering is the silver bullet in their latest book Power and Prediction. They argue that there are three core ways in which a technology can be used.

  • A point solution, which is when an existing procedure is improved upon. It can be adopted independently and doesn’t require the system within which it’s embedded to be changed.
  • An application solution, which is when a new procedure is created and adopted independently. This also doesn’t require any changes to the system in which it’s embedded.
  • A system solution, which is when existing procedures are improved, or new procedures created, by changing dependent procedures.

“The true transformation will only come when innovators turn their attention to creating system solutions,” they explain. “Those solutions themselves bring AI to an economywide scale, and their momentum spurs further application solutions.”

Productivity problems

This is a problem afflicting even the most innovative countries. For instance, raising productivity and economic growth has been at the heart of the UK government’s (disastrous) recent mini-budget, with these plans introduced despite the country being ranked as the 4th most innovative in the world in the Innovation Index. In fact, the OECD recently argued that only Russia will fare worse in terms of economic growth in 2023.

Official data reveals that while GDP grew by around 2.7% per year in the decade prior to the financial crisis, the decade since has seen growth slow to around 1.7% per year. This matters, as for all of the hoopla surrounding new technologies, they have to ultimately allow us to do more with less in order for them to be useful. This is a staggering fall from grace as the country was broadly on a par with the United States between 1997 and 2007, but the financial crisis hit the country hard. Indeed, by some estimates, it was the country’s worst performance in 250 years.

If you ask economists and policymakers why this is you’re likely to get a different answer for each person you ask, but the Innovation Index highlights how poor technology adoption is an area that deserves looking at. For instance, research from Accenture highlights that relatively few companies are currently using AI to drive competitive advantage.  The report shows that so-called AI maturity averages a score of just 36 out of 100, which highlights the long way many still have to go before lasting and meaningful value is drawn from AI investments.

“We believe every part of every business must be transformed by technology, data and AI, in some cases resulting in total enterprise reinvention,” Accenture explains. “AI Achievers are showing their peers what’s possible when you release the full potential of talent and technology, working in tandem, supported by a clear vision and commitment to change. But even this most mature group has plenty of room for growth. And while most industries have AI Achievers, they vary greatly in how AI-mature they are overall and the leaps they will make.”

This inequality is equally prominent in terms of regional productivity gaps, with large cities, such as London, performing well but other parts of the U.K. performing very poorly. Indeed, analysis from the University of Sheffield shows that Britain is more unequal than France in a whopping 15 out of the 21 measures analyzed. While there has been considerable investment in pushing the frontiers of innovation, there has been much less investment in helping those less productive firms and regions catch up with the pioneers identified by Accenture.

A global problem

This problem is certainly not confined to the U.K. as data suggests that 29 of the 30 OECD countries have experienced a similar fall in productivity. For instance, Finland is regularly touted as one of the most innovative countries in Europe, and indeed came 9th in the Global Innovation Index, but their productivity has seen similar stagnation over the past decade.

A report, called “Unequal Finland”, suggests that a regional divide is also at the heart of the productivity problems facing the country, as while the regions surrounding the capital city Helsinki tend to perform well across a range of measures, rural parts of the north and east of the country perform much worse.

These problems were also hinted at in a study produced for the Finnish government by Aalto University, which argued that the reallocation of resources has undermined productivity growth in the country. The researchers identified significant differences between both different regions of the country and different industrial sectors. They also highlight the inefficiency of “creative destruction”, with a decade of cheap credit allowing many so-called “zombie firms” to keep going when ordinarily their talent and resources would be put to more efficient use elsewhere.

The right stuff

It’s a situation that doesn’t appear to make much sense. After all, barely a day goes by when the media doesn’t report of breakthrough technologies or groundbreaking “unicorns” that are transforming society in some way. Except this transformation isn’t showing up in the data. It’s prompted many to question whether the data itself is the problem. After all, in a world in which so much of what the tech companies provide to us is free, are GDP and productivity effectively capturing the fruits of this wave of innovation?

Research from Chicago Booth explores whether the paucity of productivity improvements is more a case of measuring the wrong things, or at least the existing means of measuring productivity not effectively capturing the various technological improvements that have emerged in recent years. After all, GDP captures things we spend money on, but when we search Google or use Facebook, we’re not spending anything, so despite getting value from these services they won’t show up in the data.

The research tested this hypothesis by conducting four distinct analyses and in each found that there was no issue with our measuring methods effectively capturing what is going on in the economy. For instance, if economies had grown more reliant on the kind of technologies and innovations that have emerged in recent years, but which are hard to capture, then one would assume the productivity slowdown would be greater in countries where tech is crucial than in those where tech is not so important, yet that hasn’t occurred.

What’s more, there was also no correlation between the general importance of the IT sector to an economy and its productivity slowdown. The study also looked at the various methods for trying to get a more reliable understanding of the true value of products such as the Google search engine. On average, these studies produce a figure in the United States of around $100 billion, which while not insignificant does not account for the $3 trillion or so lost due to the slowdown in productivity. Indeed, even the highest estimate still captures less than 25% of the slowdown.

The researchers then assumed the $3 trillion in lost productivity was no longer lost. They assumed that the IT-related sectors in the U.S. generated around $1.4 trillion in value in 2015, which is higher than the $800 billion figure referenced earlier. In essence, however, we’re being asked to believe that a sector that contributed approximately 8% of GDP, in reality, is responsible for the 17% of GDP that is missing. In other words, if we’re to believe the mismeasurement hypothesis, it requires swallowing some pretty huge assumptions.

The final test involved examining the gap between gross domestic income, which measures the income of workers. Historically this figure has been higher than GDP by around 0.5% per year, or around $1 trillion. While this gap is significant, it’s unlikely to account for the slowdown in productivity as the gap was consistent even before productivity figures started to flatline. What’s more, the income growth has generally come via profits from capital rather than labor. This reflects the huge profits Google, Facebook et al are generating from the data we “pay” in return for free services.

This matters, as the productivity slowdown over the past decade, has made people roughly $9,000 a year worse off. Optimists believe that we’re at an early stage with technologies such as AI and that once we see the process-reengineering that we saw with other general-purpose technologies, such as the electric motor, then the growth engines will start up again.

“We saw the impact the business process reengineering movement had in the 90s, where similar concerns existed about the lack of impact computing was having in terms of productivity,” Soumitra Dutta, Dean of Oxford Said Business School, told me. “We’re going to need to engage in a similar kind of reinvention of the way we do things in order to start capturing the benefits of the latest generation of technology too.”

Facebooktwitterredditpinterestlinkedinmail