Robert Solow famously remarked that the effects of the IT revolution were showing up everywhere except in the productivity statistics. Other economists suggested that full exploitation of a new technology can take a long time, perhaps decades. Well, one area now clearly showing the impact of IT and automation is the labour market.
Technology is transforming the world of work, but social democrats and others appear unsure how to respond. Progressives embrace change but want technology to benefit the many and not just the few who develop, own or exploit it. Trade unions, moreover, must confront the impact of IT and automation on work as it’s the jobs and conditions of their members that are on the line. What, then, is a ‘progressive’ approach to the ‘new’ economy?
Research into the labour market impact of ‘digitalisation’ falls into three categories. The first tries to assess its impact on total employment by quantifying the number and type of jobs at risk. It has contributed to a surfeit of scare stories in the media about ‘robots taking your job’. The fear animating this is that automation and smart computers will eliminate millions of jobs, condemning people to drudgery or idleness. For example, a widely-reported 2013 forecast estimated that 35% of all occupations were at risk. But other studies that analyse jobs rather than occupations produce much lower estimates, as job roles within occupations will likely vary rather than being destroyed outright. It is also possible that job satisfaction will be enhanced by the elimination of a lot of boring tasks.
ETUI researchers investigating digitalisation accept that the world of work is changing but focus on how best to prepare. For example, the sequencing of job destruction (and creation) is critical, as the earliest innovations tend to be labour-saving rather than employment-creating, a phase which comes next as new industries emerge out of the old. Evolutionary economics can be helpful in identifying where we are in the transition to a new type of economy based on manipulation of data and sharing of productive resources – hallmarks of the knowledge economy.
We may already have passed through a turbulent ‘transition’ phase, marked by crisis and dislocation, in the development of a new economic paradigm. Now, we are moving hesitantly towards the ‘deployment’ phase. Here, new technologies may come into their own in the fight against climate change as the disruption entailed by automation provides us with a temporary window of opportunity to ‘Green’ the economy before it’s too late. The challenge for progressives is therefore to smooth the technological transition to cleaner industry while minimizing job losses.
Quality jobs and middle skills
The second area of research focuses on job quality. Here, concerns arise about how the digital transition can be fairly managed to avoid the wholesale replacement of ‘good’ jobs with ‘bad’ ones. We all know a good job when we see or do it. It is one that offers interesting, safe, and secure work at a level of compensation that permits the worker to live a decent life and participate in society. In the digital age, the definition of a bad job can be expanded to include those where humans may be subordinated to machines or lines or code, entailing a diminution of their autonomy, and quite possibly safety.
There is ample evidence of accelerating shifts in employment patterns due to the replacement of formerly well-paying factory and service jobs by robots and algorithms and the emergence of new forms of economic organisation mediating the worker-employer relationship. We are seeing a ‘hollowing out’ of the labour market whereby high and low skilled work is increasing at the expense of medium skilled work, particularly where this involves performance of routine tasks.
A baleful consequence of this is obviously a rise in inequality. But the problem goes much deeper than a change in the relative supply of good and bad jobs. Automation is moving from factories to offices as service employment gets transformed thanks to the emergence of the ‘sharing’ and ‘platform’ economies which together are radically decentralising service provision.
In the sharing economy workers compete with each directly rather than via an employer to provide a service, such as taxis (Uber). The concern to trade unions is that this undermines the traditional employer-employee relationships and creates a parallel labour market devoid of proper contracts, any possibility for organised wage bargaining and basic employment rights.
The platform economy, on the other hand, involves the brokering of work to freelancers in an online ‘marketplace’ for talent (taskrabbit, Mechanical Turk). These digital platforms colonise previously informal and non-marketised spheres of the labour market. This entails a reorganisation of activities in these industries which decomposes traditional employment relationships into vulnerable self-employment where workers are denuded of their rights.
The platform and sharing economies can be a lonely, alienating place to work. They entail a labour market characterized by, on the one hand, agile firms whose owners and managers can take economic advantage of the new technologies and, on the other, the new ‘galley slave’ class of isolated, largely freelance workers.
The third research area ponders the scope for regulation and collective action to manage the new economy and restore the balance of power between workers and the owners of the technologies. One cause for optimism is that digitalised labour markets will be more fragmented, but also more interconnected. This may challenge trade unions to find new ways of organizing and make judicious use of the new technologies in reaching un-organised workers.
The European Commission has made tentative steps towards regulating the collaborative economy and AI, although trade unions want much more to be done. Upskilling workers by investing in training may also help them find employment outside the collaborative and platform economies. But experts also urge a rethink of current education and training policies to ensure that workers’ skills complement machines rather than directly compete with them.
All is not lost for fairness at work. But let’s not rely on the benevolence of the machines and algorithms, or their masters in Silicon Valley, to deliver it for workers without being prompted.