The Future of work
This Megatrend is possibly the most complex one to analyse, and for that reason I left it at the end of this foresight exercise. Work is spread over so many areas that drawing a global view is basically impossible. There are also significant variation in different Countries and Regions but this is mostly affecting the timeline of evolution. In some areas transformation may already be ongoing and will complete within this decade, in others it has not started yet and may be far for completion by the end of the decade. However, it is mostly a matter of time (and local policies may have a significant impact). Eventually the transformation will affect all Countries, because they are all part of the same market and tick to the same economic rules.
When looking at the future of work we can do that from several points of view. A most important one is surely the one of who is doing what and in particular assessing the impact of automation (work delegated to machines). The opening graphic summarises the conclusion of a study by the US Bureau of Labor and Statistics discussed in an interesting report by Bain published in February 2018 “Labor 2030: the Collision of Demographics, Automation and Inequality“. Although the data and forecast are US specific, the factors leading to the trends are applicable in a global context (sooner or later).
Even at first glance, it is clear that the impact of automation in the 15 sectors considered results in lower wages and loss of jobs (11 red bubbles signalling a loss, versus 4 grey bubbles showing a gain). Since the surface of each bubble is proportional to the number of people involved it is also evident that many more people are going to lose than those that are going to gain.
Gains are expected in Computer and Math related jobs, and in installation and repair jobs in terms of wages but these areas will see an overall job loss, whilst the Management area is expected to see increased number of jobs but at lower wages (no areas seem to be gaining in number of jobs opportunity and in wage). We can expect a (slight) wage increase in computer and math related jobs because they require very skilled (and smart) people and this is a scarce resource. However, the number of jobs is expected to decrease as computers (and artificial intelligence) can take care of some activities. Same reasoning goes for Biz and Financial Operations. In the case of Installation, maintenance and repair we expect that although automation will be able to take care of some of the activities involved, most of them will still require actual human intervention and more and more skilled intervention, hence the expected increase of wage and the very limited decrease in jobs opportunities. As more activities increase in complexity and the processes shift from value chains to ecosystems direct human management is expected to grow but the abundance of related skill keeps wages decreasing.
In terms of job losses, it is expected that by the end of this decade some 20-25% jobs will be lost (1 out of 4-5 jobs will disappear) with low income workers being hit the hardest. The new wave of automation is based on two main pillars: Digital Transformation and Artificial Intelligence:
- the Digital Transformation moves a number of activities and processes to the cyberspace, with the consequent disappearance of jobs that were needed to run them in the physical space (think about travel agencies being hit by the on line end-user autonomous reservation and ticketing). Notice that DX is killing jobs not by substituting them with a machine but by removing them from the value chain (unlike automation on the assembly line where a robot steals the job of a blue collar, in the DX the job disappears). DX is going to hit most (in terms of jobs) the areas of healthcare and production (production already suffered from the first wave of automation with robots replacing blue collars, now it will suffer from the softwarization of production);
- the Artificial Intelligence augment machines capabilities hence making possible to replace human workers, including white collars, with machines. Administrative support jobs and sales are the ones suffering the most, as shown in the graphic. Automation in the transportation sector, construction/extraction, food preparation and service, personal care and services will require a broad mix of technologies but artificial intelligence will be the crucial enabler.
It is important to notice that the drive towards increased automation (hence higher capital investment and job losses) is fuelled by the decreasing availability of work force, in particular of skilled work force. In turns this is bound to fuel inequality since the economic benefit of automation will go, mostly, to those having capitals (the rich) and those with crucial skills in high demand (estimated in 20% of the workforce). The two biggest
economies, China and US, are already suffering from inequality and this decade is likely to increase it. Also, the speed of automation uptake can have significant impact on inequality since a fast speed will create loss of jobs without providing the time for re-qualification (historical data show that workers re-absorption rate in the US was 0.7 million per year, in this decade the expectation is a job loss around 2.5 million per year, thus generating 1.8 million people that lose their job and cannot find a replacement) and would increase the demand for investment capitals thus increasing the leverage of rich people.
The access to resources, including healthy food, preventive healthcare and good cure is clearly influenced by wealth and the rising inequality will have a significant impact on life expectancy. According to Bain forecast, see the graphic, a significant impact is deriving from the education level, and this of course correlates to inequality and fosters inequality, to the point that the education level can be taken as the measuring stick for life expectancy in young people.
Rising inequality is also likely to increase Governments intervention thus changing, in several markets the balance between private and public. We are already starting to see this happening in Europe and other areas as result from the pandemic stress on the economy.
GDP growth in the last 65 years (1950-2015, see graphic on the side) was fuelled in equal parts by workforce increase and productivity increase. In this decade the workforce (the one needed, with high skill level) is bound to decrease in several areas (like in Europe by 0.5% per year) or not be au-pair with the growth in the previous decades (in the US the expected growth for this decade is 0.4%, less than half the ones of the previous decades). To keep the GDP trend, productivity has to increase and automation is expected to result in a 30% productivity increase by the end of this decade.
more to follow