The forward march of robots halted? Automation, employment and inequality
Despite the breathless rhetoric surrounding the rise of the robots, we are not on the cusp of a ‘post-human’ economy. Automation and digitalisation will change the economy in coming decades, but, as this week’s OECD report argued, fears of imminent mass technologically-induced unemployment are overblown. By contrast, technological change under neoliberalism married to the existing distribution of economic assets in society means automation risks an imminent explosion in inequality.
Crucially, the future will not be technologically determined. The pace, direction, extent, and distributional effects of automation will be shaped by our collective choices and institutional arrangements, by how we embed technologies in broader social, economic, and political systems, and the distribution of power in society that shapes this process. In other words, the machine age will be human shaped with automation a key site of political struggle. As IPPR’s recent report on automation argued, public policy should actively shape the direction and outcomes of automation to help build a future of shared plenty.
The growing capabilities and falling cost of artificial intelligence and robotics in both physical and cognitive tasks has led to claims we are entering a transformative new machine age that will dwarf previous waves of automation in terms of the scale, speed and scope of the disruption it causes. There is something to this. The combination of rising machine performance and the falling cost of technology has the potential to drive radical productivity gains, both eliminate and reallocate employment in service sectors and professional occupations as well as in manufacturing, and reshape how production and consumption are organised. The integration of autonomous technologies will also challenge foundational assumptions about the economy: the role of employment as the primary means of distributing economic reward, labour’s position as the central factor in production, notions of scarcity, and how we organise working time, among others.
Yet despite the growing capability of machines, work is likely to be transformed by automation in the coming decades, not eliminated. Tasks are likely to be redefined, with labour focusing on tasks that are harder to automate and where humans retain comparative advantages over machines. Productivity benefits will be recirculated creating new sources of demand for employment, with labour reallocated not made redundant. The nature of employment will therefore change, but machines as yet don’t herald the death of work.
There is of course no room for complacency. Automation will likely boost labour productivity and cause a decline in the numbers of some kinds of jobs in some sectors. Managing this fairly will be a crucial challenge, particularly given the UK’s poor record in managing the shift to a post-industrial society. Yet in aggregate job losses are likely to be offset by an increase in the demand for labour in other sectors, and in other kinds of jobs. Managing automation therefore requires a focus on improving the quality and rewards of work while reducing its quantity, rather than strategies that are overly concerned about the imminent end of employment.
The critical challenge of automation then is likely to be in distribution rather than production. If the benefits are fairly shared, automation can help build an economy where justice is matched to prosperity, and where wealth, income and working time are equitably distributed. Managed poorly, automation could create a ‘paradox of plenty’: society would be far richer in aggregate as the power of AI and robotics boosts output, but, for many individuals and communities, technological change could reinforce inequalities of power and reward as the dividends of automation flow disproportionately to the owners of technologies and businesses, and the highly skilled.
Who owns and controls capital consequently becomes even more crucial in the age of automation. If automation leads to lower average wages or working hours, or loss of jobs in aggregate, a significant amount of national income could be transferred from labour to capital. Even if wages do not decline, if relative rewards to capital rise more quickly, the share of national income going to capital would increase. In a society marked by sharply unequal patterns of capital ownership this will drive rising inequality.
At the same time, low wage jobs have five times the technical potential to be automated of high paid jobs, while new technologies are already facilitating an intensive and exploitative form of “digital Taylorism”. Conversely, technological change is likely to increase the incomes and quality of work of highly skilled labour in roles which augment machines. Technological change therefore risks further polarising the labour market between ‘lovely and lousy’ jobs, while boosting capital’s share at labour’s expense.
Without reshaping the foundational institutions of our economy, technological change risks reproducing and amplifying existing inequalities. To ensure automation works for the common good, we recommend three key steps.
First, fairly realising the benefits of technological advances will require the managed acceleration of automation and the adoption of digital technologies throughout the economy. While the top 1 per cent of firms have seen average productivity growth of around 6 per cent per year since 2000, one-third of UK companies have seen no rise in productivity at all, linked to how these firms adopt and use technologies. We therefore need measures to provide greater support for firms in all sectors and parts of the country to integrate new technologies, improve management, achieve higher rates of investment, and enable a stronger voice for employees in shaping the use of technology at work. There should be a particular focus on ensuring adoption in the everyday economy – sectors such as retail, care, and transport and logistics – with policies that can draw on the ability and know-how of ordinary workers to drive the effective integration of technologies. Alongside this, the skills system also needs to be reformed to ensure people can thrive in an era of greater human-machine collaboration.
Second, new public institutions are needed to inform and regulate how automating technologies are used and to ensure that society responds proactively to the profound ethical issues raised by robotics and artificial intelligence. As the ongoing Facebook/Cambridge Analytica story underscores, we cannot leave the future rules and norms of our economy, society and democracy to be determined in advance by the tech giants. Instead, an active public must shape it, which is why we have recommended the establishment of an Authority for the Ethical Use of Robotics and Artificial Intelligence to regulate the use of automating technologies.
Third, new models of collective capital ownership are required to ensure that everyone has a claim on the dividends of technological change, to enable automation to work for the common good. We have therefore set out new models of ownership that would hold wealth in common and democratise capital at scale. These include a Citizens’ Wealth Fund that would own a broad portfolio of assets on behalf of the public and pays out a universal capital dividend and the creation of employee ownership trusts to give workers a stronger stake in the firms for which they work, and an ownership claim on the value they help create.
Taken together, these can ensure that rather than fearing automation, we can establish the institutional foundations to support a future of shared plenty.
Mathew Lawrence is a senior research fellow at IPPR. He tweets @dantonshead
He is the co-author of Managing Automation: Employment, inequality and ethics in the digital age