An economy without people

By Ken Wolff  

Last week I suggested that modern economic theory has lost sight of people but the reality is now becoming that many segments of the economy require fewer people to undertake the work and that has serious implications not just for the people losing their jobs but for the broader economy.

The loss of jobs is not new. In Australia since the 1970s there has been an ongoing loss of un-skilled jobs, particularly for males. In 2006 Sue Richardson, with the National Institute of Labour Studies at Flinders University, wrote in Unemployment in Australia:

By 2001, at every age, at least 20 per cent of men with no post-school qualification were not in the labour force. These men have not withdrawn from the workforce because they have handsome alternatives that mean they do not have to work … Overwhelmingly, the reason they are not in the labour force is because they cannot find work and have given up looking.

And in 2004 Bob Gregory wrote in Between a Rock and a Hard Place: Economic Policy and the Employment Outlook for Indigenous Australians:

An exceptional feature of the Australian labour market over the last three and a half decades has been the loss of unskilled male full-time jobs. This loss has been so substantial that as a proportion of males 15‒64 years of age one full-time job in four has disappeared. Most of this job loss has fallen upon the unskilled …

At the same time there were skills shortages. In September 2004 the Australian Industry Group reported a shortfall of between 18,000 and 21,000 in the manufacturing sector for skilled tradespersons.

And a shift in the make-up of the workforce was already occurring. In 2006, the then Department of Employment and Workplace Relations (DEWR) reported that between 2001 and 2006, 78% of new jobs were created in the four most highly skilled occupational groups: Professionals (17.4% growth), Associate Professionals (18.3% growth), Managers and Administrators (28%) and Tradespersons (10%) but job growth for Labourers had been only 0.6% in that five-year period.

Skills shortages continue. In February this year the Department of Employment published its list of occupations in which there were shortages during 2015. Twenty-five occupations were experiencing national shortages, including higher level occupations like surveyors, optometrists and audiologists. The list included trades such as motor mechanics and automotive electricians, bricklayers, glaziers, roof tilers as well as wall and floor tilers, air conditioning and refrigeration mechanics, chefs, hairdressers and cabinet makers.

The department also released its jobs outlook, Australian Jobs 2016. Overall employment was projected to grow by 8.3% over the next five years which appears to be little more than matching population growth. Six industries were expected to grow by more than ten per cent: Accommodation and Food Services (12.0%); Arts and Recreation Services (10.8%); Education and Training (13.0%); Health Care and Social Assistance (16.4% and it is also the largest industry by employment numbers); Professional, Scientific and Technical Services (14.8%); and Rental, Hiring and Real Estate Services (11.9%). Employment in Mining will decline (‒14.1%) as will Manufacturing (‒5.3%) and Agriculture, Forestry and Fishing (‒3.1%).

When it comes down to occupations ‘Professionals’ provided 41% of new jobs from 2010 to 2015 and employment is expected to grow by 14.5% to the end of 2020. Seventy-four percent of professionals hold a university degree. Professionals are also more likely than other workers to work full-time.

In that period of five years up to 2015 ‘community and personal service workers’ provided the second highest proportion of new jobs, 22% (they also make up about 10% of all employment). Their employment has grown by 16.3% since 2010 and is expected to grow by another 19% in the next five years. Nineteen percent hold a university qualification and 42% a VET Certificate III or higher but 55% are employed part-time. It includes child, aged and disability carers, waiters and bar attendants and baristas.

Technicians and trades work provided 11% of new jobs up to 2015 and employment grew by 5.2% in that time and is projected to grow 5.5% in the next five years. Most of these workers are employed full-time and 62% have a VET Certificate III or higher.

Among the lower-skilled occupations there were, late in 2015, 1.7 million clerical and administrative workers, 1.1 million sales workers, 1.1 million labourers and 740,000 machinery operators and drivers which is still about 40% of the Australian workforce. Their projected growth to the end of 2020 is 1.6%, 9.3%, ‒1.3% and 1.0% respectively. The groups include receptionists and office managers, checkout operators, real estate agents, truck drivers, forklift drivers, delivery drivers, cleaners, kitchenhands and packers, as well as labourers.

The more rapid projected expansion of highly qualified occupations appears consistent with the experience in America identified in 2013. But in America there had also been a loss of middle-ranking jobs, largely due to the automation of routine tasks, not only for manual labour (classified as routine manual work) but by the computerisation of office, sales and administrative work (classified as routine cognitive work). There had been an increase in the number of jobs for non-routine work, both cognitive and manual. The former (non-routine cognitive) requires higher levels of education and generally commands higher wages, but the latter (non-routine manual) involves work such as cleaning, food services, security services, home help, and so on. This is leading to a polarisation of the workforce in America, with more high-paid jobs, more low-paid jobs, and fewer in the middle.

The basic problem with those projections is that they are based on what has already occurred and do not take full account of the increasing pace of technological change nor the areas into which it might move in the coming decades (and the Australian projections are short-term, only for five years).

In January this year CSIRO (and Data 61) released “Tomorrow’s Digitally Enabled Workforce” in which it found that up to 44% of current Australian jobs were under threat of being replaced by robotics and other computerisation. It found that, as yet, there was no evidence of the ‘hollowing out of the middle’ in Australia but it did find:

… Australian men, particularly single men with less education, are becoming increasingly likely to drop out of the labour force. … Despite strong jobs growth in the service sector, it appears that for a growing number of men the labour market has little to offer unless they re-train.

That reflects the earlier findings of Sue Richardson and Bob Gregory. But it also found that in the future there will be a greater need for individuals to create their own jobs and even a need for higher skill sets to access entry-level positions. So a new flexible education will be required. Similarly, workplaces will need to be more flexible (which can also lead to greater casualisation and use of contract workers). Some changes, however, may lead to greater disparity in regional areas, particularly for older workers: past experience suggests that displaced older workers in regional areas do not relocate to find work but, if forced to, will relocate to cheaper housing in the same location. So new approaches to unemployment and transition to work will be required.

The difficulty with the emphasis on education and higher skills is that has already been happening in America but simply creating an oversupply. It was found that highly educated workers were being pushed down the employment ladder into lesser-skilled positions, pushing the low-skilled further down or out of the workforce altogether.

The CSIRO report did not go into the detail of individual jobs but its estimate of jobs at risk was based on a model used by Oxford University researchers who did a study of the US labour market: The Future of Employment: How susceptible are jobs to computerisation? That report found that 47% of US jobs are at high risk of computerisation. Given the work currently being undertaken on driverless cars, it is foreseen that in the next decade or two driverless trucks will become the standard form for movement of goods and many truck drivers will become redundant. Some have suggested that on major inter-state routes driverless trucks should actually have their own lane. So governments will also need to respond to those changes.

In Australia, Rio Tinto is already automating its Pilbara iron ore mines with driverless trucks and automated charge drilling and setting machines, and is also hoping to have driverless trains to deliver the ore to port (tests have been conducted but recent software glitches have delayed implementation). That is an example of even some skilled work being automated but the huge driverless trucks do require the worksite being ‘landscaped’ to suit.

As more data becomes available and can be stored, office and administrative support positions will be affected and further encroachment into manufacturing will take place. Even aspects of the construction industry could be affected as robotic prefabrication of parts takes place in factories, requiring fewer workers for the actual construction and even circumstances where robots can piece the prefabricated parts together on-site (as has already occurred in Japan).

The availability of ‘big data’ is important in expanding the reach of computerisation. For example, an American oncology centre is using computers to provide chronic care and cancer treatment diagnostics. This could be done because data from 600,000 medical evidence reports, 1.5 million patient records and clinical trials, and 2 million pages from medical journals were able to be stored and used for ‘benchmarking and pattern recognition purposes’. Examining such a vast amount of data would be impossible for a human but not a computer (if it is programmed correctly). As more ‘big data’ becomes available computers will expand their reach.

The report into the US workforce also found that automation will move into non-routine manual work in sales and services. A simple example is the increasing availability of robotic vacuum cleaners capable of replacing at least one task of a hired cleaner: how many other tasks will follow? This follows the historical pattern of technological change, namely breaking down apparently ‘skilled’ jobs into smaller unskilled components — the move from the traditional skilled ‘carriage builders’ approach to early car manufacture to Henry Ford’s production line manned mostly by unskilled workers. If this prediction proves correct, the growth in low-skilled service jobs that has been occurring in the US will actually begin to reverse during the next couple of decades.

This does not mean that all jobs in these areas will be lost but a significant proportion will be.

Jobs least affected will be those requiring creativity and social skills:

… generalist occupations requiring knowledge of human heuristics, and specialist occupations involving the development of novel ideas and artefacts, are the least susceptible to computerisation.

Others have since suggested that even some creative work can be done by computers: already there are computers capable of creating musical scores (no doubt based on ‘big data’). If that continues into the future, and AI becomes a reality, there will be almost no job that is not at risk.

This new world is giving rise to what is known as ‘the gig economy’. This means that, like a band of musicians, people will work ‘gigs’ for which they must search.

A study released in January this year by The Aspen Institute in the US found 45 million Americans (22% of adults) were working in the gig economy providing ride sharing, accommodation, food delivery and other platform-based services. For 14 million it was their main source of income and just over half, 23 million, were young, aged 18‒34.

However, most workers (72 percent) believe companies should be doing more to provide benefits, and more than two-thirds worry that as independent contractors and not employees, they don’t have a financial safety net.

Anecdotal evidence from participants suggests that many, but not all, see such ‘work’ as extra income to meet bills and so on, or as income in periods between mainstream permanent work. Or they are working many jobs to achieve a reasonable income and that can create problems — such as lack of sleep. Some examples:

I keep busy but I have to constantly juggle different gigs every day … What scares me most is that I have no guarantees, no steady pay, no stability. Everything could end overnight, so I can never make long-term plans. [young woman in Turin]

In the short term, this way of working works, but there is a long-term downside. It’s very difficult to build a future, to save for a downpayment on a house, say, or to save for [a] college fund, on a full-time Uber driver salary or even if you combine multiple freelance services. [Uber driver in Los Angeles]

I definitely advocate this way of working, but it’s not for the faint-hearted — if you’re working three or four jobs in a day, you need to be very disciplined and have a keen sense of priority. You have to be a bit of a workaholic: finding the balance and boundaries to fit everything in can be a bit of a juggle. And obviously not having paid time off is a downside. [Airbnb host, charity worker and interior designer]

Some professionals can do better in the gig economy as the internet (and specific sites) allow them access to a much wider range of clients in a much wider range of locations. For some, whose work can be done over the internet, the location of the client no longer matters and, therefore, having access to a larger number of potential clients is an advantage. Conversely, the sites involved also permit potential clients to more closely match the skills of their selected professional to the work required.

Companies will rely less on full-time employees and will hire on a task or project basis. This will apply across many job categories, not just professionals. Companies will be able to hire the specific skills required for a single task, so the work could range from a few hours to a few days. People will be able to specialise and offer their skills to many clients. But this will also spread the problems described by people currently working in the gig economy.

Such an approach is already moving down the employment ladder to very mundane tasks, and to ‘micro-tasks’ such as tagging images, extracting keywords, checking address data, which sometimes may be no more than a few minutes work for each ‘job’. These are termed human intelligence tasks (HIT):

At the time of writing [August 2015] there were about 300,000 HITs on offer on AMT [Amazon Mechanical Turk]. An average Turker (as they are referred to by AMT) can expect to earn US$2 to US$5 per hour on a good day but there’s no guarantee in terms of regular work availability.

Not all of these jobs will remain. Uber, for example, is already investing in and preparing for driverless vehicles. And computers can already undertake some of the minor tasks currently available. Whether people can be prepared in time for the new jobs that may emerge, or there will simply be massive unemployment, will be the big question.

All of the above may prove to be wrong as we do not have a very good record predicting the future — some things change less than we foresee while we seem to completely miss other significant changes. In the 1960s some popular magazines were predicting that by the early 2000s we would have flying cars, or at least hover cars. I also recall a television program from that time, specifically looking at future change, that predicted we would be required to work only ten hours per week in the new millennium to maintain our lifestyle.

Cars have changed, being more luxurious and incorporating many more safety features than in the 1960s but they still have wheels and still require roads. And, in Australia prior to the GFC, individual work hours were increasing, not decreasing. Our lifestyle has changed: the average new house is now twice the size it was (as is the cost); more homes incorporate central heating and/or air conditioning; televisions have grown both larger and smaller and few households are now satisfied with only one. But the biggest change that few, if any, predicted was the explosion in digital technology; the rise of computers, the internet and the information age; and now the portable devices that allow us to access that information at any time and almost any place.

So do we have the rise of robotics and ‘the gig economy’ right? We cannot say with certainty. But to the extent that they are already happening we do need to plan for them and consider their ramifications both for people and the economy as a whole because so much work will no longer require people, or require them for only short durations. We may not get it quite right but we cannot ignore it.

Next week I will consider Modern Monetary Theory (MMT) and what it may offer to meet the challenges of the new economy.

What do you think?

How far can robotics and computerisation go in reducing the need for humans? Is there a limit?

Will unions themselves become redundant if more and more people do not have work or enter (or are forced into) ‘the gig economy’?