Most recent graduates will know the struggle of the application process involved in trying to secure that first job. Email after email after email. Hoop after hoop after hoop. The process itself is enough to put a graduate on the back foot in the labour market. However, should graduates be more concerned with the potential for automation and computerization of the jobs they are hoping to secure? Is there really a robot right around the corner looking to deny you employment in your graduate area and what should you do about it if that’s the case?

Robot worker.jpg
The future of the smoko?

This issue has gained traction in popular media outlets recently, with many organisations attempting to evaluate the potential job losses and the industries most likely to be affected by automation. Many studies have already been done in the United States and Europe, with wildly varying results. The best method with which to measure the effects of automation is still disputed. Many studies have measured in terms of the ability to automate entire occupations.

For example, a study done by Carl Benedikt Frey and Michael A. Osborne of Oxford University in 2013 concluded that 47% of US employment were at risk of being automated. This result was astonishing – nearly half of US total employment was classed as a high risk of being automated in the “next decade or two”. They also note a strong negative correlation between wage growth and skill development and industrial automation.

A study by Deloitte in 2015 on the Swiss job market found that nearly half of the current jobs in the market faced obsolescence through automation. Of particular relevance to university graduates was the fact the authors estimated a 95% probability that accountancy would be automated, along with a 40% chance that financial advice would be automated. Comparatively, lawyers, doctors, psychologists, architects and civil engineers faced only a 1-3% probability.

In stark contrast to these similar results are the ones of an Organisation of Economic Cooperation and Development (OECD) working paper published this year. The paper begins by criticising the occupational approach of previous studies like Frey and Osborne’s, which assume that whole occupations will be automated rather than specific tasks within an occupation. They argue this will lead a significant overestimation in models. Using a task-based approach to compare to the occupation-based approach, the authors find a much lower 9% of jobs on average are under threat of automation across the 21 OECD countries. There is some variance across countries, which can be seen in this graph below.

Of particular interest is the difference in estimates for particular occupations. In Frey and Osborne, a retail salesperson faced a 92% chance of automation, but the OECD paper only finds 4% of retail salespeople do not use face to face interactions or group work in their jobs – tasks the authors logically note as difficult to automate. For accountants, the 98% figure falls to 24% when considering these task requirements.

There is also significant variance between countries, with the Slovak Republic facing a far worse situation (45% of workers in jobs at high and medium risk of automation) versus Korea (only 25% of workers in these bands).

Source: Arntz, M. T. Gregory and U. Zierahn, OECD Working Paper (2016)

The level and qualitative impact of automation remains controversial. However, there are some key facts that are agreed upon in the body of research to date. Jobs requiring social skills, creativity and group work are far less likely to be computerized. The skills which remain difficult to automate with today’s technology also align, broadly, with the ones being earmarked as desirable by firms looking for graduates in Australia.

Source: Graduate Outlook Report 2013

Both Frey and Osborne and the OECD working paper find that low wage workers are those to be most affected by automation. Frey and Osborne pointed out that automation of the twentieth century hollowed out middle class jobs with the great factory production lines of the United States and elsewhere being replaced with machines, resulting in a surge of employment in to higher and lower wage jobs. Comparatively, both papers suggest that this will shift to target predominantly low-skill, low wage jobs, and as a result, these workers will bear the brunt of the adjustment costs associated with these changes. This is a potentially problematic situation given the state of advanced economies today; many already face significant government debt, limiting their ability to invest in retraining and education for workers displaced and many are already feeling the political and societal impact of current levels of inequality (see my previous post).

Source: Frey and Osborne 2013

The graphs seen above highlight that in Frey and Osborne’s modelling, jobs with lower average median wage face a far greater probability of automation and those with a lower proportion of bachelor degree’s or higher face a greater probability of automation. This all seems to point to the need to push more of the labour force in to universities. According to the OECD, in 2010 nearly 45% of Australians aged between 25 to 34 had attained tertiary education. This is a common trend across the developed world, now being referred to as the ‘massification’ of higher education. Partly as a result of this, today graduates face a long wait for employment and an ever increasing proportion seek further education to secure future employment or take jobs outside their field of study. In 2015, Graduate Careers Australia found in their annual GradStats report that one fifth of their respondents had elected to return to full time study after finishing their bachelor’s degree. 1 out of 10 of those with a Bachelor’s degree and available for full time work reported that they were unemployed and still looking for work.

Some reports give even more dire predictions. A report published in 2015 by the Foundation for Young Australians found that for the average Australian it will take five years to secure full time employment after study, albeit this used data across school, vocational and university study respondents. In 1986, the average time to secure employment after study was just one year.

As a result, developed nations seem to face incompatible macroeconomic pressures. The labour force in recent decades has achieved a high standard of education, driven by need, societal expectation and through the effects of past automation. Despite this, the pressure to increase the height of this already high bar remains significant. Problematically, graduates already face record waits to secure employment in their relevant fields, with a weak job market, hugely qualified workforce and international competition for places all contributing to this wait. In this country, we have some of the most qualified customer service representative and bar staff of any nation in the world.

This conundrum seems to be one that is out of control of an individual graduate largely, other than potentially taking note of the fields which require higher education but are also earmarked to be affected by automation, potentially factoring this in to their choice of degree. However, for this issue to be addressed broadly, government and the private sector must find solutions.

Some academics, like the authors of the aforementioned OECD working paper, suggest that technology will not breed less jobs, but more. By driving productivity and allowing better utilization of current resources, business can reinvest these profits and thus generate more jobs, leading to a growth in opportunities. The phrase “we always create new jobs” is not uncommon to hear in debates around this issue – this seems to be an easy way to suggest an economic mantra of belief in the market. Even progressive proponents like tech entrepreneur Nick Hanauer suggests that we do not need to panic about job creation, as humans have always found work to do, citing the industrial revolution and the birth of computers as times that we have adjusted and created new jobs to fill the void.

However, others are not so optimistic. Scott Santens, a writer and journalist, has been a key proponent of a universal basic income, suggesting that it is all-at-once the solution to the problems of inequality and the threat of automation. The Basic Income Earth Network (BIEN) states that a basic income is “an income unconditionally granted to all on an individual basis, without means test or work   Santens argues that automation has led to the creation of many ‘bullshit jobs’. These jobs he characterizes as very acute in their specialty nature, with little room to acquire new skills and little room for employee development. He questions the need for these jobs altogether, suggesting that if these are the jobs that we are creating in the place of already low-skilled jobs in the future, should we not to turn to a system which ensures basic living standards?

Prominent economists like John Kenneth Galbraith, Milton Friedman, James Tobin, James Buchanan, James Meade, Robert Solow, Paul Samuelson and F. A. Hayek have all supported the idea of a basic income in some form. Widely read philosophers like Bertrand Russell and Henry George have also written of their support for the idea in previous centuries. Various US political leaders of different times have given their voice in support of the idea, including Franklin Delano Roosevelt,  Martin Luther King Jr., Lyndon B. Johnson, Richard Nixon and Jimmy Carter. More recently, Switzerland voted down the idea, 75% voting no and 25% voting yes. The current, though under-siege, opposition leader Jeremy Corbyn has suggested he will investigate and consider the proposal of a basic income if elected British Prime Minister – however, this remains seemingly unlikely at the moment. In the United States, Bernie Sanders expressed sympathy to the cause as well.

Galbraith (left), a structuralist economist, and Friedman (right), a monetary economist, both supported different types of basic income guarantees

This is not the only solution being put forward, with the idea of a federal job guarantee being supported by modern monetary theorists (MMT) such as Bard College economist and professor Pavlina Tcherneva and Australian economist and Newcastle University professor Bill Mitchell. Effectively, this social program would mean the government acts as an employer of last resort, ensuring that in times of poor economic growth those who become unemployed do not remain so. In this way, full employment is achieved and spending on welfare benefits is reduced by putting those unemployed in to government jobs. Another MMT economist, L. Randall Wray, argues that this program would cost 1 – 2 percent of gross domestic product in the United States. Interestingly, Tcherneva is critical of the basic income as threatening to the value of the currency and macro stability, whilst being a covert way for conservatives to remove the currently established social safety net. Wray and Tcherneva compare the two policy options in a 2005 paper published in the Rutgers Journal of Law & Urban Policy, concluding that a combination of the two policies would be best, utilizing a job guarantee and a guaranteed basic income for the young, aged and disabled.

Pavlina Tcherneva (left) and L. Randall Wray (right)

Both of these programs require significant federal funding, something which governments lack the political will to institute and, as many mainstream economists argue, simply cannot afford (of course, MMT economists reject this argument). However, experiments are being carried out in the private, not-for-profit sector with basic income guarantees funded through donations and crowd funding. One such organization is GiveDirectly. They provide a direct cash transfer to poor communities in Kenya and other African nations, researching the societal and economic impacts of the guaranteed income using a set that do not receive the payment to compare to those who do not. Early reports of results are promising, showing an effective investment of the transfer and a significant increase in consumption as well as production in the populations given the transfer. Other experiments are being run in many countries by government, particularly in Scandinavia where the major Utrecht basic income experiment will begin soon. Despite this private and public endeavour, it seems implementing either of these strategies at a federal government level remains a distant dream in the major developed nations even in the most progressive of nations. However, Argentina has rolled out a version of the jobs guarantee, with mixed results. The Bolsa Familia in Brazil is nearly 13 years old, which gives small cash transfers to poor families – in ten years the program was credited with helping to halve the country’s extreme poverty. The country’s Gini coefficient – the widely used measure of income inequality – also improved in that time. Bolsa Familia was serving one quarter of the population by 2013.

Graduates should be aware of computerization and automation. For the broader labour market, low-skilled and low-wage workers should be acutely aware. The utilization of robots in our production processes is not a new phenomenon, and estimates of the effect are wide ranging – from dystopian to utopian. However, it remains clear that the majority of jobs requiring social interaction and group work seem safe bets for graduates, as artificial intelligence as a technology to be utilized for automation is still some time away. Although in future decades, developments in this area may change the face of automation. Radical solutions available to government include the basic income guarantee and jobs guarantee, which proponents argue address the potential loss of jobs and increase in wealth and income inequality. Neither, however, have the necessary political support to gain legislative approval in the developed world at this time. Awareness, not fear, seems the graduate’s best choice, as well as being prepared to wait to find full time employment whilst economists and governments agree on how to address this growing issue.






Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s