“In the U.K., one in five jobs is going to be automated… technology… has a contradictory element because on the one hand it makes you more precarious as a worker. On the other hand, it shows what you can be when liberated from work because you’ve got all this extra time. You can imagine different ways of living. You can pursue your passions. You can live a happy life.” Ash Sarkar, Teen Vogue, July 15th 2018
‘Post-work’ has become increasingly fashionable among left-wing commentators and represents a radical response to the widely assumed consequences of automation. Post-work supporters argue that the advances in robotics and artificial intelligence mean that there is no longer enough paid-work to go around. However, where many see robots stealing our jobs as a dystopian nightmare, the proponents of post-work maintain that this is not necessarily a bad thing. They point out that many people have little attachment to their jobs anyway and, therefore, automation has the potential to set us free from the drudgery of bad jobs. In fact, automation could actually be a cause for further rejoicing, for it is argued to herald nothing less than the very end of capitalism, thus ushering in a new era of post-capitalism, perhaps even ‘fully automated luxury communism’.
The above claims are, however, based upon fallacies. Whether automation eliminates jobs or creates jobs is dependent on whether demand is elastic. Automation can increase efficiency, thus making society richer and, at the same time, increasing the quality of what is produced. If better quality products are available, and there is more wealth with which to buy these products, demand will increase and more jobs will be created. This is why, despite 98% of the tasks required for cotton production having been automated, the number of weaving jobs has actually increased since the 19th century – not to mention the associated rise of retail, distribution, design and marketing jobs. Elastic demand isn’t a general law but it does seem to be what is currently taking place in many sectors. Moreover, the fact that certain tasks can be automated does not mean that entire jobs can be eliminated. Jobs are made up of a heterogeneous mixture of very different tasks, many of which remain difficult to automate.
The advocates of post-work often counter that economic theory and historical facts are no guide to the future as, unlike past automation, it is now highly skilled tasks which are at risk. But, as sociologist Colin Crouch points out, this argument relies on a false understanding of both skill and history. For example, Crouch highlights that, for thousands of years, the highest skilled task undertaken by an architect was their ad hoc ability to calculate the stresses and strains of materials. Yet, this is now a routine task carried out by computer software. This has not, however, led to the elimination of the ‘architect’ as an occupation but it has enabled those in this profession to move their focus to other tasks instead. Overall, what the above adds up to in the foreseeable future is the probable elimination of some jobs. The OECD estimates around 9% are at high risk, while many more new types of paid-work are likely to be created. However, the suffering and insecurity caused by shunting people between old and new jobs will nevertheless be significant.
A further fallacy is that automation undermines the basis of capitalism. In a nutshell, this argument relies upon adopting a partial and misleading reading of Marx and his theory of value in order to argue that labour exploitation is no longer a necessary condition for profit and thus the ability of capital to capture value has been fatally undermined. These arguments have been powerfully critiqued elsewhere. Therefore, I will focus on a different fallacy, one which has been less discussed but is of more immediate political relevance. This is the idea that ‘post-work’ would be welcomed by those exploited under the current wage system.
In a recent book chapter written by myself and Cambridge sociologist Brendan Burchell, we review the literature regarding the impact of employment and unemployment on well-being. What is clear from the empirical evidence is that employment can cause great harm. In particular, the insecurity to which many workers are exposed can lead to stress and anxiety. This insecurity can take many forms, such as threats to the continuation of employment, job role, or other valued job features such as working hours. Given the suffering which employment causes many people, should unemployment then be welcomed as a source of relief? Unfortunately, there is also overwhelming evidence, including longitudinal research, that unemployment also causes both poor mental and physical health for most people. Moreover, this suffering is not only the result of financial hardship due to loss of income.
Based on ethnographic research, social psychologist Marie Jahoda argued that unemployment caused psychological distress due to the fact that employment provides us not only with an income but also five ‘latent’ categories of experience. And it does so regardless of whether we are aware of the benefits of these experiences. For example, when someone becomes unemployed they might actually miss social contact with someone they previously hated interacting with. Jahoda’s latent categories of experience are: time structure, externally generated activity, social contact, collective purpose, and status and identity. While hard to test empirically, there is a body of quantitative, qualitative and longitudinal literature which supports Jahoda’s framework. This means that alternatives to paid work, such as social security benefits or a universal basic income, can only go so far in replacing employment.
So why is it that employment seemingly provides us with these beneficial latent categories of experience? Jahoda is often accused of adopting a teleology in which employment is seen as fulfilling essential human needs. Indeed, she does at times deploy such arguments. However, at other points, she suggests a more sociological argument based on her view of employment as “a central institution in capitalism”. This is an important insight, as sociologist Erik Olin Wright argues that institutions which “systematically impose harms on people require vigorous mechanisms of active social reproduction in order to be sustained over time.” This suggests that the institution of employment, an institution which is at the very heart of capitalism, and has existed for around 400 years, must be actively socially reproduced. According to this reading of Jahoda the social reproduction of employment is based upon the daily conditioning which results from our engagement with common social institutions such as school, the family, market exchange and employment itself. For example, Jahoda argued that school and family “impress on the young the value of punctuality and the need to fill the day with planned activities” and that collective purpose, status and identity are shaped by the lived experience of a market society in which we learn to measure and value ourselves against others according to levels of income.
If the psychological distress of unemployment only exists due to social conditioning, does this not therefore open up the possibility to overturn these relations and instead create post-work values? Jahoda cautions against such optimism, as most people are unable to “single-handedly overthrow the compelling social norms under which we all live”.
In fact, during the 1930s, Jahoda undertook an ethnography of a ‘post-work’ Quaker cooperative community made up of unemployed Welsh miners, featured in David Fryer and Philip Ullah’s collection Unemployed People: Social and Psychological Perspectives. The Quakers provided the unemployed miners with the resources to produce a wide variety of goods but the miners were prohibited by the Quakers from using what they produced for market exchange. Unfortunately, Jahoda reports that the cooperative did little to stop the sense of dislocation which these unemployed workers experienced and illustrates how this failure was due to the coop’s for-use-production jarring with the miners’ internalisation of capitalist values. Despite being ardent socialists, the miners had also been socialised since childhood to measure their comparative status according to their level of income, and to see their life purpose in such terms. This is not to deny human agency but Jahoda warns us that “the tendency to shape one’s life from inside-out operates within possibilities and constraints of social arrangements… in this sense we are both active and passive [at the same time].” In other words, we can make our own history but not under conditions of our own choosing.
There is then a ‘chicken and egg’ problem: while capitalism continues to exist, the institutions which socially reproduce the wage relationship will compel most people towards needing employment in order to meet their psychological needs regardless of the quantity of jobs on offer. Therefore, the destruction of paid work by automation would, even if this were clearly contributing to the end of capitalism, cause mass social suffering. Post-work is thus unlikely to carry much political support among those whose jobs are threatened. A more promising direction would be to focus on supporting workers to organise for the improvement of both existing jobs and the jobs of the future. As Marx once argued, it is only in struggle that workers unite through the formation of organisations and associations and, thus, become aware of how their political interests diverge from capitalism. A post-capitalist politics must, therefore, be grounded in the struggle against ‘really existing automation’ and the insecurity, polarization, inequality and work intensification which it threatens.
Alex J. Wood is a Researcher at the Oxford Internet Institute