ESA National Economic Panel Polls
ROBOTS, ARTIFICIAL INTELLIGENCE AND THE 'FUTURE OF WORK'_Jeff Borland
By Jeff Borland, Professor of Economics, The University of Melbourne
The impact of technological change on the Australian labour market and income distribution should be relatively limited, according to respondents to this poll.
The proposition that rising use of robots and AI is likely to lead to a substantial number of workers being unemployed for longer periods was disagreed with by a majority of respondents (58%). Mostly, this was on the basis of a belief that while new technologies destroy jobs, they will also create an approximately equal number of new jobs. The history of new technologies having this balanced effect on job destruction and job creation was also emphasised.
Respondents who were uncertain or agreed (42%) do not seem to have expressed opinions that were too far from those who disagreed. Some recognised that new technologies cause job creation as well as job destruction, but did not believe it was possible to be confident that there would be sufficient creation to offset the destruction. Others focused on the direction in the question to ‘hold labour market institutions and training fixed’, and argued that this restriction would impede the scope for job creation with the new technologies. This is an important point, as there is no doubt that the minimal effect of past episodes of major technological change on employment has been because the workforce had adapted its skills (for example, through increased education).
The proposition that new technologies would bring sufficient gains to allow workers who are negatively affected to be compensated brought forth an even larger majority – in this case agreeing (73%). In fact, a reading of the comments suggests that support for the proposition was almost unanimous. Most of those who disagreed with the proposition seem to have done so because they did not believe that compensation would actually be provided, rather than because it would not be feasible to provide compensation. But this was a viewpoint that was also shared by many of those who supported the proposition.
It is interesting to compare the responses by Australian economists to those made by US economists who were asked (essentially) the same questions by a recent IGM poll. On the second proposition, economists from Australia and the US are in substantial agreement – with 88% in the US agreeing and 73% in Australia. Reponses to the first proposition, however, differed significantly – with US economists being more likely than their Australian colleagues to think that new technologies will cause longer periods of unemployment. In Australia, the shares who disagreed, were uncertain and who agreed were respectively 58%, 24% and 18%. Corresponding shares in the US were 21%, 29% and 38%. In part, the cross-country difference seems to have been due to a larger proportion of US economists putting weight on needing to hold labour market institutions and training fixed – and expressing the view that this would constrain the extent of employment creation. Perhaps the difference also reflects a more optimistic outlook by economists in Australia, not having recently been through an experience such as the Great Recession.