Terminator Economics? The impact of Automation on the Northern Ireland economy?
There is nothing new under the sun. In 1817 machines were being smashed because it was feared they destroyed jobs and economists debated whether mechanization was a good thing. In 2017 one still hears the “Lump of Labour fallacy”; the pessimistic view that there is only a fixed demand for work, so robots and Artificial Intelligence (AI) will inevitably reduce employment.
In 2014 Deloitte and two Oxford academics, B. Frey and M. Osborne argued that about one-third of UK jobs were threatened by automation over the next 20 years; particularly work in office and administrative support, sales and transport. And those with wages less than £30,000 were five times more likely to be automated than those earning more than £100,000.
Such gloomy headlines need to be qualified. Deloitte and Frey and Osborne said “threat” and not actual change. Their figures related to a 20 year period- not a single blow. Other researchers (Arntz, Gregory and Zierahn at the Organisation for Economic Co-operation and Development/OECD in 2016) took a different approach and said that we should ask instead what job tasks as opposed to broad occupational areas were liable to be automated. The OECD research suggested that only about 10% of jobs were threatened by automation.
In contrast to the old- Luddite type fears about machines and robots, there is a more optimistic view whereby machines become the delivers of wealth and prosperity. Automation should improve productivity. Higher productivity should produce lower prices and hence in turn lead to higher consumption and output.
In a report in 2017, PwC estimated that automation could grow the Northern Ireland economy by 5.4% by 2030; compared to 8.4% in Scotland, 9.8% in Wales and 10.6% in England.
However, just as we should beware of the Lump of Labour Fallacy, there could be another fallacy in this line of thinking. Automation certainly increases the productive potential of the economy (the maximum amount of goods and services which can be supplied) but it does not guarantee there will enough purchasing power to buy up all those extra goods and services. Some see automation as the ultimate Keynesian nightmare (See note 1); machines have wonderful productive potentials but they end up standing idle because so few people are left earning enough to buy what could be produced. I think this nightmare scenario is exaggerated but we should be cautious.
Clearly, this is a controversial area. In making predictions we are subject to three difficulties:
- Prediction is difficult because it is about the future!
- People tend to focus on what is happening to them as individuals- if a robot takes your job it is little consolation that society is better off.
- The most important implications of automation and AI in the long run may lie outside of economics; e.g. what are the implications for political freedom, how wars might be fought and human identity given the application of computer implants?.
Some conclusions about the impact of automation on the Northern Ireland economy:
- Beware of fallacies. Society does not have a fixed lump of necessary labour. Equally, just because machines can produce more there is no guarantee our economy is organized in such a way that there will be a buying power to demand what is produced.
- The “threat” is greater in Northern Ireland than the rest of the UK- this is because of a relatively high representation of low wage, low skill and simple task work.
- Conversely, the opportunity presented may be smaller here than in the rest of the UK. This is because we start from a lower base where the percentage representation of high productivity/high technology sectors is lower.
- Brexit and a likely reduction in the scope businesses have to import cheap labour from the EU could influence the extent to which automation continues. In a number of sectors, Northern Ireland businesses could be presented with the choice either draw on the local supplies of unemployed and inactive labour OR use more machinery.
- For certain, the rise of the machines implies the need for skills in the economy will increase (See note 2).
- Keynesian, after the economist J.M. Keynes who argued the Great Depression of 1929-32 was caused by insufficient demand (purchasing power) in the economy. Writing at the end of the 1920s Keynes foresaw a point by the end of the 20th Century at which productivity would have grown so far that we would have average working weeks of only 15 hours. This has obviously not happened! I think today’s predictions of technologically induced mass unemployment are similarly exaggerated. These fears do, however, exercise some. The computing entrepreneur E. Musk as argued that Western governments need to introduce an adequate basic income payment for all citizens regardless of whether they are in work or out of work (because of robots). This idea of a Universal Basic Income is the subject of an ongoing experiment in Finland and is being considered by the Scottish government.
- Overall skills needs relative to supplies are analysed in the Ulster University Economic Policy Centre 2017, Northern Ireland Skills Barometer 2017: Skills in Demand.