17 May 2019

Short-termism risks clouding the profound challenges of automation

By Lucky Dube

Technology has proved to be a source of a great deal of good in society: it has facilitated greater interconnectedness, it has contributed to making all sorts of things cheaper and more accessible, and it has provided people with greater access to information. Society has been and will continue to be changed by it, and as such its long terms effects should be a subject of continual interest for policymakers.

To the extent that technology – particularly technology replacing tasks currently done by people – could adversely affect society, minds should concentrate on how negative impacts can be mitigated. The trend towards job polarisation – where the share of jobs that pay a moderate wage as a proportion of all jobs is in decline, and the share of low and high paying jobs is increasing – and the displacement of people by technology will pose many challenges to society in the long term. They are challenges that will require a well thought out public policy response that goes beyond technophobic fearmongering.

Automation involves replacing tasks currently done by workers with technology, which could include computer programmes, algorithms and robots. PwC estimates the proportion of jobs in transportation and storage, manufacturing, and construction at risk of being automated to be 52 per cent, 45 per cent, and 23 per cent respectively. The proportion of jobs in health and social work, and education sectors are estimated to be at least risk with 18 and eight per cent of jobs respectively potentially being automated.

The same analysis found that jobs occupied by the moderately educated and those with little education were at high risk of being automated (35 per cent and 45 per cent respectively). Jobs occupied by those with the highest education level were least likely to be at risk of automation, with 12 per cent of jobs at risk.

This could reflect greater adaptability to technological change, and the likelihood of such people occupying senior roles that require the application of human judgement. In any case, many people (especially those with some or very little education) could find themselves displaced by technology and little prospect of finding well paying work in the industry they have experience in, and they could face barriers to retraining to find better paying work in another industry.

We tend to think that automation could result in higher unemployment, as displaced workers are left competing for the few jobs that haven’t been automated. Such a view demonstrates what economists sometimes call the ‘lump of labour’ fallacy, where there is a misconception that the total amount of work is fixed and can be distributed to create more or fewer jobs.

The fallacy is present in populist arguments against immigration (more working age immigrants will result in fewer jobs for native born workers) and the policy of Herbert Hoover to reduce the working week, in the hope that the ‘remaining’ work would be made available for the unemployed. When applied to technology, the view does not take account of how technological change could make products cheaper and more accessible, which would increase demand for them and perhaps increase the overall quantum of work.

In his book The Third Pillar Raghuram Rajan, the former Governor of the Reserve Bank of India, argues that “the direct effect of technological change may not be the aggregate quantum of human work but its redistribution”. He discusses how the highly skilled tax lawyer could see her earnings increase, as well as the work she has to do, with the middle class tax accountant being worse off, and the entry-level assistant not having much hope for additional skilling or career progression.

Economists Maarten Goos, Alan Manning, and Anna Salomons observed this kind of job polarisation across Western Europe between 1993 and 2010. The proportion of middling jobs – office clerks, machine operators and assemblers – as a share of total jobs fell by 9.27 per cent. In the same period the proportion of high paying occupations – corporate managers, health professionals, and engineering professionals – increased by 5.62 per cent, and low paying occupations – labourers, salespeople and demonstrators – increased by 3.65 per cent.

In the UK the fall in the share of middling professional jobs was more marked at 10.94 per cent, with the share of lowest and highest paying occupations increasing by 4.17 per cent and 6.77 per cent respectively. Without alternative routes into highly paid professions, aspirant workers may find themselves stuck in low-paid, low skilled professions, which could entrench a feeling that ‘the system is rigged’ and play into the hands of populists offering fake solutions.

The long-term issues highlighted pose several challenges for policymakers, not least how to offer alternative routes into high paying professions, and retrain those that have been displaced and would like to work in a different industry (especially for those seeking a change in the middle of their working life).

Such reforms would have to take care in making education and training more accessible, whether through degree apprenticeships, greater provision for distance learning, and ensuring that people can access information to learn about available opportunities. It could also include making the regions outside London and the south-east most exposed to automation more attractive to businesses by improving transport links, and encouraging more innovation activity in HE institutions.

This could go some way to counteract the feeling that places beyond the capital and the Home Counties have been ‘left behind’ or ‘forgotten about.’

It took the EU referendum in the UK and a surge in support for populist causes around the world for discussions about how globalisation has affected society to be prioritised.

Constant focus on short-term problems prevented policymakers thinking about how policy could lessen the negative impacts of globalisation: those negative impacts have driven the rise in the appetite for populism. If that constant focus on the short term persists, something similar could happen with technological change.

CapX depends on the generosity of its readers. If you value what we do, please consider making a donation.

Lucky Dube is a research intern at the Centre for Policy Studies