In The Guardian, Wednesday 22^nd March, 2017, U.S. economist Robert Shiller argues for the taxation of robots on the grounds that they are a ‘labor-displacing innovation’ that will lead to job losses.

Acknowledging that ‘retraining programs for displaced workers’ may be essential public policy, Shiller sensibly goes on to invoke the human and community importance of maintaining paid work.

However, we tend to disagree with the proposition that the use of robots should be taxed in order to restrict job losses in particular market sectors.

In the fluid, internationally competitive globalised 21^st century economy, structural changes to the job market are driven by market forces, and attempting to intervene with those market forces to artificially stop job losses in a particular sector, while it may provide short-term personal income security and vocational continuity for the workers at risk of redundancy, is unfortunately a recipe for longer-term economic damage to the national economy that takes such measures.

Among the market forces at work in today’s globalised economy is the internationally competitive drive to produce products and services as efficiently as possible. The more efficient producers of the same products and services will tend to succeed in the international marketplace, while the less efficient ones will fail because of their need to charge higher prices to meet the higher costs of production, or if they cannot be profitable at the lower prices set by the competition.

For over 500 years, advances in automation, from printing presses replacing the laborious hand-written reproduction of manuscripts, through automated telephone exchanges replacing the previous manually operated switchboards, to continuously editable computerised databases replacing hand-typed documents, have continually driven up the efficiency of production in business.

The economic effects of increased efficiency of production are mostly positive ones. These include lower consumer prices for each product or service as a proportion of average income, thereby bringing more services and products within the reach of each individual. They also include reduced working hours to create the same output; and, where manual labour is concerned, a trend towards less harsh physical labour.

Robotisation is a relatively recent chapter in this long-standing trend of using machinery and technology to drive up the efficiencies of production and lower the costs of goods and services. But it is nonetheless comparable, and we think it will be similar in its economic effects to previous advances in automation.

Where automation leads to a reduced need for workers in a particular market sector, the money saved by industry on salaries will be retained in and ultimately cycled back within the economy as expenditure on other products or services. In economics parlance, this is called the conservation of a constant total value of economic resources per head of population, such as average spending power and available worker hours.

More money will be retained by consumers of the products and services whose production has been automated, as a result of reduced purchase costs for those products and services, leaving those consumers more money to spend on other products and services; or it will be retained by the owners or shareholders of the businesses producing them and recycled partly through greater government tax receipts from personal incomes and business profits, and partly through greater expenditure and investment by the beneficiaries of higher incomes and profits. Where the saved money ends up being redistributed within the economy, there is the potential for new jobs to be created, replacing those that have been lost.

Shiller also quotes Edmund S Phelps in according great personal importance to the ‘calling’ of the individual. Yet the notion that everyone should be able to choose where and how to work in response to personal vocation is economically unrealistic because the supply of willing labour for certain types of employment exceeds the demand for labour in these fields. In the fields of entertainment and creative arts, where the number of willing producers and performers of music, performing arts, fine art and literature exceeds the market capacity, many aspiring  musicians, writers, artists and other performers cannot make a living from their calling. Those market forces are essentially similar to the market forces at work when jobs are lost in particular sectors and job opportunities created in others. In the free-market economy, the onus is on labour to adapt to the opportunities available, and not on industry to adapt to the desires of labour for jobs in particular areas whether or not there is money available to make those jobs viable. Demand dictates where labour opportunities are available.

Shiller’s desire to reduce income inequality is laudable. But selectively taxing robots, which are to a greater or lesser degree a part of the means of production in certain industrial sectors only, is not an equitable means to this end, and we doubt it would be an economically effective one.

It is not equitable because automation that increases the efficiency of production and reduces the need for labour to achieve a given level of production does not consist exclusively or even mainly in the use of robots. In fact, automation exists in degrees on a continuous spectrum from hand-operated weaving machines serving as aids to the efficiency of clothing producers, through to fully automated production line assemblies, with computerised data flows and telecommunications also serving to automate communications that would previously have required a great deal more labour. It would be impossible to meaningfully and reliably quantify the amount of labour saved by automation of all kinds; and selectively taxing only certain types of automation responding to narrowly defined parameters would be arbitrary and lead to economic injustice, with limits being set to the forms of automation that are being taxed based on emotion.

It is unlikely to be economically effective because it is very hard to conceive of the implementation of a global international consensus on the taxation of robots. Unlike environmental policy, for which there is a well-established framework of international co-operation and agreement, tax policy remains a matter for national or regional supranational governments. If robots are taxed only in the UK, or only in the EU, or only in the USA, for example, but not in South East Asia, the businesses operating in the territories that have taxed their use will be put at a disadvantage in the international marketplace, their prices undercut by those operating in countries that do not tax the use of robots. As a consequence, jobs in the industrial sectors where robots are in use will in any case be lost in the countries where robots have been taxed. Taxing the means of production could spell disaster for international competitiveness.

Another reason why it is unlikely to be economically effective in the countries where it is implemented is that it will be slowing down the economic development of those countries by artificially propping up the labour market in certain industrial sectors or companies that are no longer viable, at the expense of job creation in other areas, and at the expense of overall economic growth and prosperity.

There are tried and tested means to address economic inequality that do not involve the introduction of economic distortions and inequities such as those that would result from a selective tax on robots or other means of production. For example, the use of properly calibrated progressive personal income tax rates with a tax-free personal allowance; a zero-tolerance policy towards corporate tax evasion and offshore tax havens; fair national living wage regulations; and a social security net to protect those excluded from adequately remunerative employment by market forces. It is ultimately up to each country to decide where the right economic balance lies in all those areas and to legislate accordingly.

This decision-making process belongs to the domain of politics. But whatever decisions are taken should be applied fairly and impartially across the board. Singling out certain arbitrarily delimited means of production such as forms of automation meeting a particular definition of ‘robots’ for special penalties would be retrogressive, not progressive, from the standpoint of the desire to build a fairer society.