by Andrew White | July 6, 2017 | Comments Off on The Role of Digital and its Impact on the IT Productivity Paradox
I read with great interest a paper last weekend: The Coming Productivity Boom: Transforming the Physical Economy with Information, Michael Mandel and Bret Swanson. It is now marked up all over the place which explains the level of involvement I had in agreeing and disagreeing with the authors, and the resulting follow-up with additional reading.
The authors suggest an “information gap” exists and that closing this gap would help return productivity to levels not seen for some time that would in-turn drive GDP growth significantly – sufficiently to paper over many of the current economic woes we all observe.
The authors suggest:
- Digital industries, which include what other economists would call IT-producing industries along with publishing, digital content and services (those that deal in information such as insurance and finance) have seen much higher productivity than other physical industries (such as manufacturing, retail, healthcare etc.) over recent years.
- There is plenty of IT spending all around but it is not necessarily being spent in the right things (somewhat analogous to Robert Solow’s famous 1987 quip).
- How IT can change all industries (e.g. through business innovation) is the key, not just ‘spending on computers’.
The authors are quite close to the truth with their paper, as I see it. But they don’t quite go far enough.
The facts are these:
- Growth is achieved by working more hours and/or increasing productivity
- Labor hours are hugely dependent on working population which is flat or declining, and population growth (influenced by birth rate and immigration) generally and (declining) participation rate (those among the working age that chose to work).
- Productivity is driven by changing the input/output ratio of work. This is where IT comes in.
- Productivity has gone through several cycles of growth, roughly:
- Before WWII (quite low),
- Somewhere between 1950s and 1970s (comparatively high),
- Post 1990s to date (low again or even slowing).
- There are arguments about the exact duration of each cycle but the low-high-low sequence seems generally agreed.
- IT has been shown to be a driver of previous high levels of productivity growth. IT-producing industries tend to lead with highest growth. IT-using industries (in excess of 15% capital investment is for IT) second and non-IT-using least. There was high correlation between IT capital investment (e.g. capital deepening) and improved productivity, for a while.
- This all changed since somewhere in the late 1990s. Despite the high levels of IT spend, productivity for most industries slowed and has since been anemic- low, flat or even falling. And this is across most developed nations.
- Clearly there is a misallocation of funds at a local level (firm’s investment decisions) that results or shows up in a reduction in total factor productivity at the macro level. Just as Chad Jones suggests (see The Facts of Economic Growth, page 23, which is an excellent and fascinating paper in its own right).
So what is this misallocation?
We need to understand how IT drives productivity to be able to understand the answers to this question, and to uncover the decisions that need to be changed in order to drive new and higher productivity.
And if you thought, as some have, that there is a measurement problem with productivity and GDP, try David Byrne, John Fernald, and Marshall Reinsdorf. They work the numbers and show even if there is an issue it or far too small to explain away the scale of productivity collapse. See “Does the United States Have a Productivity Slowdown or Measurement Problem“.
Mr Mandel and Mr Swanson develop some interesting ideas and present examples for how IT might drive productivity across a range of physical industries including transportation, education, healthcare and energy. Overall the authors explore possibilities; they give examples how certain technologies might drive improved output for less input. What they don’t do is offer a framework to help guide either the public policy maker or the CEO/CFO/CIO to determine how they should change behavior with respect to IT investment. It is as if the acquisition of the right kind IT is enough to yield the results. There seems little focus on how IT should be use or how behavior needs to change.
The transportation example is the second most developed argument explored in the paper. By the end of the analysis you will conclude that yes, productivity will increase, but it’s a game of swings and roundabouts and it is hard to determine to whom the net productivity improvement will accrue. Correctly the authors argue that any analysis needs to look at the entire ecosystem, from travelers and consumers that need transportation, taxi services, ride-sharing services, vehicle ownership, vehicle production, autonomous driving, as well as the financial arrangements spanning taxation and profits. As such it is not clear who wins and loses.
The healthcare industry is the more comprehensively analyzed industry example. I get the feeling that the authors have done more consulting in that industry. The authors throw almost every possible IT driven improvement at all aspects of healthcare. And yes, it does seem that industry is ripe for improvement. But alas I find a gigantic gap between the potential detailed in the paper and how the healthcare ecosystem behaves. There is a gulf between the two.
So what is needed are not generalities about how IT might bring improvements. We have had those for years, indeed, since before the 1970s when IT started to drive improvements. What we need is a framework that can explain what has happened to industries that did exploit IT; and that can be used to predict how to drive additional improvements through changed behavior aided by IT, not delivered by IT. Such a model will than guide policy and decision makers. Then we have to hope that rational minds prevail.
Read Complimentary Relevant Research
The Future of Work and Talent: Culture, Diversity, Technology
The future of work is about forging a new relationship between technology and talent that transforms existing ways of working and doing...
View Relevant Webinars
How to Create a Data Strategy for Machine Learning Initiatives
THIS WEBINAR HAS BEEN CANCELLED. WE APOLOGIZE FOR THE INCONVENIENCE. Machine learning initiatives demand a data strategy that is comprehensive...
Comments or opinions expressed on this blog are those of the individual contributors only, and do not necessarily represent the views of Gartner, Inc. or its management. Readers may copy and redistribute blog postings on other blogs, or otherwise for private, non-commercial or journalistic purposes, with attribution to Gartner. This content may not be used for any other purposes in any other formats or media. The content on this blog is provided on an "as-is" basis. Gartner shall not be liable for any damages whatsoever arising out of the content or use of this blog.