BEYOND4.0 was invited in April 2019 to participate in a workshop of the TECHNEQUALITY-project. TECHNEQUALITY is one of the two other projects that receives funding from H2020 in the same Call as BEYOND4.0. The PLUS-project is the third project. TECHNEQUALITY looks at the same topics of digital transformation as BEYOND4.0, but has a different focus. The issues of equality and how it is/will be impacted by the digital transformation stands central in this project. Also, the research team is more focused on the use of econometric tools to understand the changes. Some of the differences with BEYOND4.0 came to light in the discussions during the workshop.
One of the central presentation was of course that one by Carl Benedict Frey (Oxford University) on skills and digital transformation. He started with his well-known 47% of computerisation degree of jobs. He complained that even the New York Times only reduced his work to this percentage. During his presentation, he defended his approach to underpin this 47%, commenting on a great number of other studies on calculation of computerisation degrees. He then pointed to his current work that will be published in a new book on the technology revolution. I hadn’t really realised the background of his work. His seminal article of 2013 (2017) was thoroughly embedded in econometric calculations. But in fact, he is not that much focused on econometrics (that is work for his co-author Michael Osborne): he is more focused on understanding economic history. If I understand this correctly, this explains for me that the 47% is only a figure for the very long term. There is no underpinning for short or middle term impacts. The question then arises with me, what in fact the policy consequences can be? In this sense, his thinking is mainly focused on understanding the potential of technology on the long-term. We need methodologies to understand the shorter and medium term if we want to have policy relevance. Anyway, that is my opinion.
From this perspective, the presentation by Glenda Quintini (OECD) is an effort to make the calculations of Frey & Osborne more realistic. I find that everybody is trying to grapple with the 47%, and finding different ways to optimize the calculations. However, as long as we mainly focused on the potential of technology, I am afraid that these efforts may be little relevant for policy makers. An example may make clear what I mean. Computer Integrated Manufacturing (CIM) was a technology development widely discussed in the 1990s. The potential of human-free factories (‘at last!’) was seen as a realistic future. If you check the literature, you will see that CIM has completely disappeared from recent publications. Potential does not mean reality. The main improvements to the methodology of Frey & Osborne are probably not really advancements. One of the core changes to their approach, is the task-focus in the analyses. I have my doubts that this will help. As long as we are thinking about the potential of technology, tasks, jobs, distributions etc. will not add any real value to the calculations.
In this discussion, the presentations by Pascale LeBlanc and Hannah Berkers (Eindhoven University) did add to the progress in thinking. They looked at how collaborative robots impact work of logistic employees. Repetitive jobs are reduced to left-over jobs. In major warehouses operated by Dutch platform companies, remaining tasks were limited to those activities that robots could not handle. Unexpectedly, employees were not happy of the fact that their tedious repetitive jobs were disappearing. The presenters discussed at length the ethical ramifications of these technological impacts. The sociological approach made the changes in the workplace much more tangible. Their presentation was also helpful for my own presentation. I wanted to point out that we tend to reduce the impact of technology on job content, to a deterministic relationship. My observation is, based on new data of the Netherlands Employment Survey (NEA) and Netherlands Skills Survey (NSS), that the organisational context determines which skills and job quality will be found in companies. Technology may be used by management to push certain types of organisational approaches. My next observation is that, in the Dutch industry anyway, managers were directing technology to support more tayloristic work organisation. This makes me less optimistic about the chances for employees. Deskilling is a certain risk.
Anyway, both projects, TECHNEQUALITY and BEYOND4.0, have their work cut out for them in the coming years. It is great that we could start with this collaboration between the two projects. In June, we will have TECHNEQUALITY and the third H2020-project PLUS at our Summer School in San Sebastian. A pity that Carl Benedict Frey can’t make it to this meeting. Anyway, I hope I comforted him during the event with his remark that the 47% is haunting him. I pointed out that in 1979, Clive Jenkins produced a book on the potential impact of microelectronics. Jenkins estimated that 55% of jobs in the UK would have disappeared by 2003. In reality, the UK labour market had grown by more than 90% over the period 1979-2003. Nobody remembers the percentage of Jenkins.