Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Select Committee on the Future of Work and Workers

Go To First Hit

HEALY, Dr Stephen, Chief Investigator, Shifting Cultures of Manufacturing in Australia Research Project, University of Western Sydney

HENDERSON, Mr Troy, PhD candidate, University of Sydney

McNEILL, Dr Joanne, Research Project Manager, Shifting Cultures of Manufacturing in Australia Research Project, University of Western Sydney; Visiting Fellow, Faculty of Law, University of New South Wales

MORGAN, Professor Bronwen, Platform Cooperativism Australia Working Group and UNSW; Professor of Law, University of New South Wales

NEHME» , Dr «Marina» , Senior Lecturer, Faculty of Law, University of New South Wales

SHARP, Mr Darren, Convenor, Platform Cooperativism Australia Working Group

SPIES-BUTCHER, Dr Ben, Macquarie University

TONKINWISE, Professor Cameron, Platform Cooperativism Australia Working Group and UNSW; Professor of Design, Art and Design, University of New South Wales

WALSH, Professor Toby, Professor of Artificial Intelligence, University of New South Wales

Evidence from Mr Sharp was taken via teleconference—

Committee met at 08:32

CHAIR ( Senator Watt ): I now declare open this hearing of the Senate Select on the Future of Work and Workers. I welcome all of you here today. Before the committee starts taking evidence, I remind all witnesses that in giving evidence to the committee, they are protected by parliamentary privilege. It is unlawful for anyone to threaten or disadvantage to witness on account of evidence given to a committee. Such action may be treated by the Senate as a contempt. It is also a contempt to give false or misleading evidence to a committee. The committee generally prefers evidence to be given in public but, under the Senate's resolutions, witnesses have the right to request to be heard in private session. If a witness objects to answering the question, that witness should state the ground upon which the objection is taken and the committee will determine whether it will insist on an answer, having regard to the ground on which it is claimed. If the committee determines to insist on an answer, a witness may request that the answer be given in camera. Such a request may, of course, also be made at any other time. I would be very surprised if it happens today.

I now welcome representatives from the Shifting Cultures of Manufacturing in Australia research project, Platform Cooperativism Australia Working Group and UNSW, Professor Toby Walsh, Dr «Marina» «Nehme» , Dr Ben Spies-Butcher and Mr Troy Henderson. I understand that information on parliamentary privilege and the protection of witnesses and evidence has been provided to each of you. I invite each of you to make a short opening statement, and, as I mentioned, given the numbers of people, can we try and keep them fairly tight. At the conclusion of your remarks I'll invite members of the committee to ask questions.

Prof. Morgan : This is a statement on behalf of UNSW and the Platform Cooperativism Australia Working Group, so we won't all speak. Thank you very much, and it's good to be with you today. Our submission is about the importance of platform cooperativism, and it's in the context of the framework for work in today's economy increasingly being shaped by the delivery of goods and services through technology platforms.

Platform cooperatives are a new breed of enterprise that blends the technologies of the sharing economy with co-ownership models and democratic governance. We like to describe them as bringing the rules of formal cooperatives to the sharing economy and the tools of digital platforms to cooperatives, and blending those two things.

The overarching key point of our submission is that platform co-ops provide a positive and complementary alternative to the dominant platform economy. They're a creative economic strategy that is, in essence, an innovative business model. We suggest it is one that would have longer staying power and more flexibility than traditional top-down regulatory responses. By fostering and furthering collective economic agency, they help to bridge fraught and often polarised debates between the top-heavy regulatory state on the one hand and arguments in favour of free-market entrepreneurship on the other.

The critical characteristic that differentiates platform cooperatives from mainstream platforms is worker agency. The claim of mainstream platforms to facilitate micro-entrepreneurship needs to be interrogated in the light of the difference that a member ownership stake makes when they have an ownership stake in the platform themselves. It redresses the situation of the value generated by the platform being held by a very small number of people. Owner membership drives more of that value to individuals and also creates the potential for greater contribution to broader social and economic development. We're not arguing this as an either/or proposition, and our main message is that the government has a key role to play in fostering and enabling ecosystems to encourage the formation of platform cooperatives, and ensuring that they're not locked out through regulatory or policy settings, some of which actually create unintended biases towards traditional forms of enterprise.

In short, platform cooperatives provide a distinctive and innovative avenue for extending ownership by and the voices of ordinary people in the changing world of work, and this ethical slice of the platform economy is worth nurturing for that reason.

Dr Healy : Good morning. Our submission is informed by a three-year research project on the cultures of manufacturing in Australia, which was supported by the Australian Research Council. The project was formulated in response to an increasing debate over whether there is a future for manufacturing in this country. Our project has involved interacting with 10 different manufacturing enterprises operating up and down the east coast of Australia across various sectors. We've been interested in the role that manufacturing plays in providing good jobs for a range of people, but, not only that, we've been exploring, using case studies, the role manufacturing can play in creating a more just society that shares newly generated wealth and a more environmentally sustainable future. Our research thus far suggests that manufacturing has much to contribute to both these agendas, and that working with new technologies to improve the process of making is crucial to their success.

Three key points have emerged from our research thus far. First, well-run social enterprises have the capacity to expand work opportunities for those at a distance from the labour market. They can play a crucial role as intermediaries in the manufacturing supply chain, forging connections in a circular no-waste economy. For example, the mattress recycling social enterprise Soft Landing offers employment opportunities to marginalise populations and diverts a growing proportion of the 1.6 million mattresses thrown out annually from landfill. In a recently expanded partnership to reverse manufacture mattresses with a private company, TIC Group, they recover the material in all six states and territories, and have been a key player in brokering a voluntary product stewardship agreement with all players in the market.

Second: in countries across the globe, the manufacturing sector is the lead employer of the clean-economy jobs. In Australia, there is huge scope for jobs growth in enterprises that take environmental care and repair seriously. For example, the Australian branch of the international company Interface Carpets leads the way in forging a new manufacturing agenda under the rubric of 'Climate Take Back', shifting production of synthetic carpet tiles away from fossil fuel inputs and steadily increasing the proportion of recycled and reused materials in their production system.

Third: longstanding regionally-based companies committed to local communities over generations are using high-tech solutions to stay competitive. Here are two examples. The farmer-owned cooperative Norco manages three manufacturing plants that use Australian dairy to produce products for local and international markets. The commitment to farm viability has meant a very careful and lean approach to manufacturing operations, logistics around expansion and capitalisation. The Norco management team see themselves as stewards of an industry commonwealth that, in turn, contributes to the regional vitality of Australia. The second example is the Varley Group, based in Newcastle, which has a strong commitment to its regional location and to providing employment opportunities for groups who are at a distance from the labour market. At Varley we find, for instance, young people who are working to fit out ambulances that will be used in Australia and overseas, amongst other products.

In sum, our research suggests that the future of work will be improved by supporting a diverse range of enterprise types that provide goods that are not easily offshored, that create a context for integrating people into meaningful work and that contribute the conditions for ecological repair. Thank you.

Prof. Walsh : It's very nice to be able to speak to this committee. As a professor of artificial intelligence, I try and inject a bit of realism into some of the discussions about where technology—especially things like AI—may be taking us. The first piece of advice I have is to treat with caution some of the predictions being made about how many jobs are actually at risk of automation. The grandaddy of some of these concerns is the Frey and Osborne report that came from the University of Oxford back in 2013. This predicted that 47 per cent of jobs in the United States were at risk of automation in the next two decades. Similar numbers have been made for Australia and other economies. You should treat these with immense caution. I'm a scientist, so I looked at the data. Some of the predictions are very suspicious: 'There is a 98 per cent probability that modelling is going to be automated in the next two decades.' Well, I can assure the committee that none of my colleagues are trying to make robots walk in high heels down a catwalk. Even if we could—which, I imagine, would be quite an engineering challenge—there's no market for it. We're not interested in seeing what robots look like in clothes; we want to see what humans look like in clothes.

Many of these studies don't take into account many of the necessary factors: the cost of the replacement; the desirability and social acceptability of the replacement. So there's a huge amount of uncertainty in the number of jobs that will be automated. But I think it's fair to take as a conclusion from many of these different studies that we are facing a period of disruption, and that there will be some jobs—taxi driver; truck driver—and some of the more middle-class professions that will have parts of them at least automated, and we do have to think about how we prepare for this period of change.

One thing I think is clear: that we don't know how many jobs will be created by new technologies and how many jobs will be created by the increasing prosperity that new technologies bring. Maybe there will be fewer jobs. Maybe there will be more jobs. We do not know. But one thing is absolutely clear: that those jobs will require different skills from the typical skill portfolio that people need for jobs today. And that's going to require policy, that's going to require support for lifelong learning and it's going to require support for teaching students at school and at universities different skills from the ones that they are taught today. I commend to this committee the work and the thought leadership that the New South Wales Department of Education are doing to begin that conversation. The secretary, Mark Scott, I think has been very good at picking up the idea that we do have to think about: 'How do we change the curriculum? How do we support people? How do we give them the skills for the many different careers that many of them will have throughout their working lives, when the world will be quite different?' The world will be quite different in 50 years time. The world will be quite different in 20 years time.

It's worth thinking about this period of disruption; it's worth thinking of the challenges we face. We do have one good historical precedent: the Industrial Revolution. The nature of work went through a very radical change. Employment did increase as a consequence of the Industrial Revolution. Employment levels are now historically low and population levels are historically high. We did manage to create new jobs, but it wasn't without pain and it wasn't without change. We made some significant societal changes. We changed the nature of education, we introduced welfare and we introduced unions and labour laws to ensure that all of us shared the prosperity that technological change brought, and I suspect we will have to think about similar, possibly even radial, changes to ensure that all of us do receive the benefits of technology. It's also worth remarking that the Industrial Revolution didn't happen without pain. There were around 50 years of pain before the quality of life of your average worker did actually increase to what it was before the Industrial Revolution. Now, it's much greater, so I expect that we will have to go through a period of change and a period of pain again.

Finally, I would like to warn this committee about the risks from competition from overseas. Many of the changes will happen in digital markets where the barriers to entry are not great and where we will see significant competition from companies overseas, many of whom do not seem to wish to pay tax in this country. Many countries have seized this opportunity. Many countries have started to invest significantly in these sorts of technologies. It's worth pointing out one particular stand-out, which is China. China is betting its future on artificial intelligence and quantum technologies. They have said that they will have economic and military superiority against the rest of the world in the next 10 years, and they're already making remarkable progress in doing that. It is actually quite surprising, in my field, to see how China has gone from almost nothing to now being, under some metrics, the leading nation. In 2013, I was one of the people responsible for taking the leading AI conference to China for the very first time. We saw great potential for China in this area. Last year, we had the leading AI conference in Australia, which was a show of respect for the standing of AI in Australia. China was the largest nation present. For the first time ever it wasn't the United States; it was China. In fact, the number of submissions from China equalled the number of submissions from the United States and Europe added together. We put up signs for the conference in Chinese because, as I said to the other organisers, Chinese is actually the predominate language spoken by our delegates today, not English anymore.

There is a real threat posed by countries like China who are investing in a major way. They have many natural advantages. They have much more relaxed attitudes to data and privacy that will give them an edge. They have three-quarters of a billion smartphones. They are the biggest smartphone market in the world. That's going to give them access to huge amounts of data. The only competitors to Google, Facebook, Twitter and Amazon are Chinese companies because they made sure that that was the case. They have a plan. They are going to execute that plan. They are already, in many senses, leading the way in AI and some of these other technologies. We should be very concerned about the threat that that poses to business here in Australia. Many other countries, including the United States, the United Kingdom, Canada and even India, have AI plans in place and are investing. India is investing half a billion dollars in its AI plan, announced just a few week ago.

One piece of advice I have for this committee is that Australia should be thinking about what its plan is to ensure that it benefits. We do punch well above our weight in this field. We are clearly in the top 10 in the world. We have a very long history. We have a very healthy academic community. We have a very healthy start-up community; it could do with greater support from the government. Equally, theses are issues that don't just touch on the future of work; they touch probably every issue of government. I can't think of any aspect of government and it's delivery of services that won't be affected—education, taxation, almost every aspect of our society will be touched by AI. So, we really need to think about how it's going to change, not work, but change everything and, for that, I suggest you really want to think about a breadth-of-government report—like those that have been produced in the US, like those that have been produced in the UK, like those that have been produced in China and Canada, and in India recently—to think about how we are going to change our society so that we prosper and all Australians share the prosperity that these technologies will bring.

Dr Spies-Butcher : I might speak on behalf of Troy and myself. Our focus is on the income support system and the welfare system, and I think it brings up a piece on the kind of approach that the other speakers here have as well. In terms of the future of work, we're responding to technological change, which we don't see ending jobs but, rather, more transitions in the types of employment that people have over their life course—and transitions between employment and education and training—social changes which mean there's less emphasis on full-time work, more of an integration between part-time work and forms of unpaid work, and economic changes that exacerbate income inequalities.

I'll focus on the income support system, which we see as being currently unresponsive, particularly the payments like Newstart and youth allowance, which are focused on people who are expected to be in work and education. Because the income support system is so administratively difficult to deal with, quite a large chunk of workers simply don't bother with it. It's so targeted that even those with fairly limited means can't access it for significant periods of time, so it doesn't help them to smooth out their income during periods when they don't have as much or any work. The tapers are so steep that it creates substantial disincentives to forms of part-time work, and that form of targeting has now left it stigmatised so that many people who might be able to access it try to avoid accessing it. It doesn't actually do its job.

We explored how the principles of basic income—which has become very popular around the world and is currently been experimented on in a number of countries—might help us think through how we could reform this. We acknowledge that there are two large objections to basic income type proposals, particularly in Australia. One is their large fiscal cost, and the other is a kind of normative objection that Australians don't like to see high income earners receiving payments from government. We were suggesting two broad ways forward. One is that we could embrace a principle of universality, not necessarily by paying everyone the same cash payment but by relaxing the withdrawal rates that apply to current payments in ways that start to integrate with the tax system and create a relatively smooth and slowly increasing effective marginal tax rate for workers. This sits somewhere between what is normally called a universal basic income and what would normally be called a negative income tax, which try to achieve similar goals—one primarily through the tax system and one primarily through the welfare system. We think that this reflects reforms that have already taken place in, for example, family benefits in Australia, which are effectively an amalgam of a tax payment and a welfare payment and are designed to be integrated into effective marginal tax rates and into work incentives. By the way, our initial modelling—we're only at the early stages of that—suggests that you could roll it out for the entire work-age population from 18 to the retirement age for less than the cost of the current housing and superannuation tax concessions, and at a rate above the current Newstart rate. But we also think even that's a very radical change to the way that tax and welfare systems work.

In terms of experimentation—and there are lots of models of trials going on around the world—we would like to suggest that trials of randomly selected participants might not be as useful as incremental changes in all policy settings that move us towards the kind of future we want, and allow us to see whether it's working or not and adjust along the way. To do that, we're particularly interested in how we might adjust the current age pension into a universal payment—it seems to be the thing in the current welfare system that is closest to it—or one that is at least affluence tested, in the language that we use. That is, it gradually phases out in a much more steady way without the large 50 per cent reduction that it currently faces. And, given the pressures that I mentioned at the beginning, we might also look at how we might introduce this for young people—who face the largest number of these transitions between work and education and between different kinds of jobs, and quite quickly—by introducing, for example, a trial for 20 to 25-year-olds that did adopt this kind of model of affluence-tested basic income. This would allow us over a period of time to see how those people behave and also how people, like late teenagers who are expecting to be able to receive this payment, might adjust their own behaviour if they knew that they had some kind of real economic security underneath them when they were starting to engage in questions of further education and work. That's where we'll leave it. Thanks very much for having us here today.

CHAIR: Thank you very much.

Dr «Nehme» : Thank you for this opportunity to appear in front of this committee. This review is very timely in view of the technological developments that are taking place in Australia and around the world. One key consideration is the need to review the role that corporations play in the future of work and workers, and this will require a re-evaluation of our laws, including the Corporations Act. One key question is whether our corporation structure can promote innovative businesses while protecting workers. For example, social enterprises are becoming more and more popular, and they are a space where new forms of work can be found. However, the current legislation can hamper the promotion of these forms of enterprises. A rethink of directors' duty is needed, by broadening the range of people directors owe a duty to. Currently, they have a duty solely to their shareholders. Representation on boards of employees is another thing to consider, because this would create another layer of accountability for directors.

Additionally, it is important to ensure reporting transparency by providing details of how workers are employed and paid and what career development may be available to them. All these things will provide protection to workers. For example, this Monday, National Australia Bank said it was cutting 6,000 jobs over three years to invest in new technologies. They will be hiring 2,000 people with technological skills. The question that is raised by this is: why isn't NAB reskilling a range of employees that they already have? This will promote loyalties in people and help organisations build a culture to retain staff, rather than just having this turnover of staff every five years. All these things are issues to be addressed to protect workers as the future of work is changing.

CHAIR: Thank you very much. I think we've captured each of the individuals or organisations. As occurred yesterday, there's an awful lot in this. Why don't we start with you, Dr «Nehme» . I've been thinking for some time about the issues you raised around directors' duties. The focus solely on the needs of shareholders might not adequately take into account the range of interests that society has. I think I saw in your submission that you were specifically talking about broadening directors' duties to also take account of employee needs and possibly environmental needs?

Dr «Nehme» : Yes.

CHAIR: Any other sorts of things?

Dr «Nehme» : Yes, a broader range of stakeholders. It will depend on each organisation and what it aims to achieve, really. But a broader view, rather than just focusing on shareholders, is needed, because currently, yes, we can say that sometimes the law does allow directors to be flexible and look at other stakeholders, but this is all done to suit the bottom line. The company has to benefit from it at the end of the day, so there is no responsibility for a director to look more broadly than just the shareholders, and I think it is time for us to look at that—at least, the employees of the company. This is something the UK has proposed in its current reforms.

CHAIR: That was going to be my next question. Is there anywhere that's already doing this or thinking about it?

Dr «Nehme» : The UK has put a proposal to broaden its directors' responsibilities, especially in public companies, to others, including employees and other stakeholders, so there is a proposal there.

CHAIR: Is it a government review or something like that?

Dr «Nehme» : Yes, into the workers' protections.

CHAIR: You've put forward social enterprises as alternative models, and similar points were made around cooperatives, mutuals and things like that by some of the other speakers. It seems to me there's quite a tradition of those sorts of enterprises in the UK and possibly elsewhere. I've seen most about it in the UK. Is there any reason you can think of, whether it be cultural or otherwise, why that model has taken off in some countries and just doesn't really seem to have taken off to the same degree in Australia?

Dr «Nehme» : I think one of the reasons would be that we're being too conservative in the way we are looking at corporations and viewing the different types of corporations that we need. We're not looking at the broader picture and making reforms to the Corporations Act. What we're doing with the Corporations Act is just a patch-up here and there. What we really need is a review of the legislation to check if the current structures that we have are actually suitable, rather than just patching them up and creating different things. I think it is time to introduce a new form of corporation to promote social enterprises, which, as you mentioned, other countries—the US and the UK—have done, and Australia hasn't done that.

Dr McNeill : I will also add to that two things. I don't think that there isn't a culture of that in Australia, and I'd refer the committee to the recent report Finding Australia's social enterprise sector, which was produced by a collaboration led by the Centre for Social Impact, at Swinburne—our colleague there. That actually documents the size and shape of the Australian social enterprise sector. Similarly, a very recent project has just come out from just Victoria. It documents the even greater extent, once you can start to drill down into getting that level of mapping, of social enterprises in that state.

In terms of the culture and other things that may enable or inhibit—I think one of the key issues in Australia in social enterprise—that has been an area of my practice and research for probably over 15 years now. I've certainly looked at the UK situation in quite a lot of detail, and I think what we don't have here is enabling policy frameworks, particularly around things like procurement and how we use our procurement dollars from the public sector as well as from other parts of our industry to support and enable both the establishment and the growth and sustainability of those types of enterprises, which are producing outcomes for communities and for society that go well beyond just the shareholder financial return.

CHAIR: We heard a little bit about what you can do with procurement to this end in Brisbane yesterday, as well.

Dr McNeill : Good. I'm very pleased to hear that.

CHAIR: Professor Walsh?

Prof. Walsh : I was just going to inject another historical note. If you look at the modern corporation cooperatives and mutual societies, they are largely a product of the Industrial Revolution in the United Kingdom. They were there and put in place in part to protect the workers. We may have to think about—this is a revolution with a different flavour—whether we need to reinvent these sorts of systems to deal with this change.

CHAIR: Have you got a view on whether that sort of cooperative model is appropriate for the new age that we're going into, or is there something different that you think we should be—

Prof. Walsh : I think we should think about how we can encourage it, because certainly one of the trends that we see is that inequality is increasing within our society, and technology, if left to its own course, will amplify that further, so we do have to think of ways that the prosperity needs to be shared more greatly; otherwise we will see increasing societal disquiet.

Prof. Morgan : Can I just add also that, in terms of the reasons why it might not, at least on the cooperative end, have been taken up in Australia to date as much, I agree with Dr McNeill about the policy framework being more of an issue than the bottom-up culture, but the federal structure of Australian company law is also quite a challenge, particularly in that area, because cooperatives have been registered at state level and not at federal level. I would say there was a strong push in the late nineties—I wasn't living here, but my understanding is that the Keating government began a strong push—towards preferring the use of a standard form of corporation at the federal level, and that's taken a long time. I think there are some creative strategies that could be taken to mitigate that, which I think the Business Council of Co-operatives and Mutuals might well speak to tomorrow—as well as having already contributed to a former Senate inquiry on that front. But I think that working across state and federal levels will be an important pathway for encouraging this kind of enabling policy framework. And Victoria, I think, is pointing the way there.

CHAIR: Dr Spies-Butcher, you mentioned two of the primary concerns that are given as to why we can't have a UBI: cost and the Australian objection to giving rich people more money—a tradition I'm a fine upholder of! But the other objection I've heard, including yesterday, and particularly from the union movement, is that the risk is that, if we accept that we're entering a world where people aren't going to have enough work and we need systems to compensate them for that, that's giving up on the main fight, which should be finding decent, secure work for people. What's your response to that?

Dr Spies-Butcher : Absolutely, I wouldn't see those as opposed to each other at all. I suppose the way we framed this was not by saying that there's going to be an end to jobs. We agree that technology is unlikely to see an end to paid work. But it is changing the way that paid work works, and the job of unemployment insurance, we might say, in the welfare state, of protecting people from the vagaries of how much money they're going to get from a paid job in any given week or month certainly has changed. It's partly changed because we expect workers to move between occupations and jobs. Workers themselves want to be able to balance part-time work, care responsibilities, leisure and education, and the current system we have just doesn't work at all. One of the reasons it doesn't work is that it's really hard to get Newstart. If you've got $6,000 or $7,000 in the bank, you're not eligible for it, for a long enough period of time that you've probably got another job by then anyway.

So we wanted to make this payment work with the new system of work. I think that, if you had a system where workers knew that they at least had some guaranteed income, they would have a much stronger hand at work, with increased voice, and I think it would assist with the other job of being able to create employer-employee voice and power in the workplace.

This is absolutely not a proposal to do that. In fact, I think at a macroeconomic level it would do the opposite. It would act as a regional stimulus. If you just look at who is likely to receive it, this would not only reduce inequality; it would spread income outside the major capitals and lead to more jobs, I think.

Mr Henderson : Could I just add to that that New Zealand, which has moved to the more universal pension model, has higher labour force participation rates than Australia does, so we see the structure, that interaction that Ben talked about between the taper rate—so the withdrawal rates of the benefit—and the marginal tax rates to create an effective tax rate that encourages work rather than discourages work.

Prof. Morgan : And could I just add that there's a debate within the platform economy debates about creating portable benefits that is complementary to that. The line of argument that I think bridges the objections you raised—and I don't know what the unions would think of this, but the kind of portability and greater universality that Dr Spies-Butcher was talking about makes it more possible for groups of people to get together and create their own enterprises, and there are some really interesting things. There is an experiment about to happen in California with a collaboration of a member-owned platform that's going to work with the Californian community colleges association, which is roughly equivalent to TAFE, but it's an enormous system of education parallel to that of universities, with more than two million students. They're going to work together and have those young people experiment with forming enterprises themselves through this platform which supports that, as part of their curriculum, which is actually blending education and employment transition at that level.

These kinds of income support systems outlined by you are very consistent with that. It's a very different image of the relationship between work, employment and income support and education than we currently have, which is much more staged and segmented, in that you do all of the education in one bit and then you transition into a job and you stay there for a much longer time than is now typical.

CHAIR: I've got some questions of other witnesses, but do you want to pick up on the UBI one?

Senator STEELE-JOHN: Yes. We heard yesterday, from a couple of the people who attended the hearings, about the value of combining something like a UBI with a jobs guarantee. First of all, I just want to get your view on that kind of combination of those two policies.

Dr Spies-Butcher : Absolutely. Again, I don't think they're in any way counterposed. I think they're entirely complementary. I think there is a very valuable role for the government in being able to guarantee access to work for those who are available for work. I think there's a question about what form that work takes. Is that work always full time? Does it involve part-time opportunities as well? And what kind of work is it in terms of white-collar or blue-collar or levels of education? But there are clearly lots of things that need to be done in the community which currently aren't being done, and there are people who want to be able to do socially meaningful work, so that's absolutely useful. I think this kind of model would integrate with that to allow people to move in and out of work, creating incentives for work and also creating a kind of portable floor to move around between opportunities where they see them.

Senator STEELE-JOHN: I know there are a lot of trials currently underway in places like Finland, and they're not complete yet, but are you aware of any of the initial findings or experiences of the trials that are currently underway?

Dr Spies-Butcher : Do you want to talk to this?

Mr Henderson : Yes. The Finland trial created a lot of publicity, but it's very limited in scope and scale. It's taking 2,000 people from a particular cohort of unemployed people and giving them access to the basic unemployment benefit for two years, but you can earn any income on top of that without losing the unemployment benefit. It's a year in. There's no really meaningful empirical evidence yet, but it's also got a very narrow focus on looking at the labour supply response: will unemployed people take on more work if you have the guaranteed unemployment benefit?

I can just cite the far larger trials that were carried out in North America, in Canada and the US, in the 1960s and 1970s, negative-income-tax-model trials, which showed a modest reduction in labour supply, which was mainly a delay in women re-entering the workforce after having children or younger people staying longer in education. So Robert Solow, a Nobel-prize-winning economist, said, yes, there was a labour supply effect, but nothing to jeopardise the labour supply of the US economy.

Senator STEELE-JOHN: I'm also curious about this. I'm aware that in the drugs policy space there's often a conversation about the dangers of pursuing certain pathways in a vacuum. Have you got a view on the implementation of the UBI-like mechanisms in the absence of things like a well-functioning health and housing system and the ways in which that might skew results of certain trials? Where I'm going with this is that these systems seem to be working well in places like northern Europe, for instance, which have a very different existing social safety net beyond welfare than exists here in Australia. Do you have any thoughts on that?

Dr Spies-Butcher : Yes. I think there are two points to be made. One is that, if this payment is used to effectively cash out the existing universal service provision, it's going to be completely unaffordable. You can already see in the United States that the cost of health care compared to the cost in virtually any other industrialised country means that for it to keep pace with the cost of living of most people would be fiscally unsupportable. But I think that in Australia, where we have a kind of version of universal health care, the main other challenge here is what's happening at the bottom end of the housing market, where we really do have a broken housing system. It would help a lot of those people in the low end of the rental market, but it would probably also push up rents in the absence of any other housing policies that were designed to offset an effect. So, yes, absolutely we need to be thinking about how those policies interact.

Senator STEELE-JOHN: So it's important to consider these things within a broader context of reforms to other Australian aspects of—

Dr Spies-Butcher : Yes. To the extent that you make a universal service provision effective, you make a universal basic income cheaper. We very much see universal basic income as part of a policy package. It is definitely not a panacea.

Senator STEELE-JOHN: But that's slightly what concerns me about some of the discourse from some places on the UBI—that it's almost like, 'We've introduced this one thing, and, in the absence of anything else, it would fix all things.'

Dr Spies-Butcher : It's not magic dust.

Senator STEELE-JOHN: No. Could I just quickly turn to Professor Walsh, with your background in AI. We've talked a lot about automation within manufacturing, driving, transport and things like that. I'm wondering: in areas where we don't traditionally think of it, such as accountancy, law and areas like that, non-routine non-manual work, do you see a role for algorithms—I'm thinking particularly here of algorithms—or things like that in rapidly automating areas of work such as that?

Prof. Walsh : A hundred per cent. With respect to accountants and lawyers, a lot that they do is actually quite systematic and regular; it's not actually that out of the ordinary. So a lot of those aspects of those jobs are being increasingly—knowledge discovery is something that you can do. That does raise the question, though: are the law firms going to continue to hire lots of clerks at the bottom? How are people going to get into the bottom rung of the profession if those jobs are the ones that are most easily automated? There are a lot of companies looking to provide more automation into the legal space and into accounting—into all these sorts of areas. So, yes, it's hard to think of areas that will be completely safe.

There are skills, I think, that machines are very limited at today. They have very limited emotional intelligence, so anything that's very people facing is very safe. A good example, I think, is health care. I don't think we're going to have any shortage of doctors anytime soon, but some of the highly skilled jobs—I wouldn't encourage any of my children to spend lots of time becoming a radiologist. That's a task where we're going to get a machine to work out how to read X-rays much more quickly, much more cheaply, much better and much more accurately than a human can do, so that very skilled job is probably going to disappear. But we're going to have no shortage of need for a GP. At the end of the day, I think we want to have a human doctor tell us the bad news and help us interpret the results of all those machines and algorithms, so I don't think doctors have to worry, but I think probably we'll end up with many more GPs. We'll be focusing much more on preventative health care than we do today. So there will be no shortage of jobs for doctors, but there will be a different type of doctor as it goes on.

CHAIR: Do you have something to add there?

Prof. Tonkinwise : Can I just make a comment in relation to algorithms not so much replacing particular professions but coming down to being one of the systems of managing employees. Platform economies are situations in which the matching of supply and demand is being done by, quite often, a black-box algorithm, so this goes back to issues of what somebody who might be on the board of directors of a corporation can be responsible for. We're seeing evidence already, I think, of companies who plead ignorance as to what their algorithms are actually doing when they match certain advertising with certain markets. They're just allowing this to happen automatically.

I think this is an interesting issue to look at, not only in terms of algorithms replacing jobs but in terms of algorithms actually becoming the management of employees. I think this is particularly important, also just recognising that a lot of people in gig economies, people who are being pushed off the books of being employees and are therefore not covered by those protections, are being managed by interfaces. They are being managed by apps. They're not being managed by people.

I also just want to draw the committee's attention to the fact that a lot of the people making those decisions about those apps are designers, which is an extraordinarily under-researched and undereducated—I would suggest—profession in regard to labour practices and what these entail. So I think it's very important just to note that these algorithms are not only replacing work but also managing work, and these then create a whole other set of liabilities and responsibilities.

CHAIR: Can I pick up on that, and others might have something to add on this. One of the things we heard yesterday was about the need for a greater ethical framework around the use of AI in general, and the example you've given us is similar. Would you like to say anything more about the need for ethical frameworks and, again, whether you've seen examples of that being done elsewhere?

Prof. Tonkinwise : This is very much at the moment the issue du jour in terms of a number of people who've profited from these platforms now issuing mea culpas to indicate that they wish there was a little more ethics involved. Certainly as an educator I think there are lots of questions to ask about the ways in which some professions like design, which don't have any type of legal framework governing the practice, are allowed to be educated without that kind of ethical consideration, without any education in relation to, as I said, labour laws and the history of those kinds of practices. So I would certainly advocate very strongly that I would love to see the government make the education of designers and engineers certainly spend more time considering the consequences, both speculative and real, in terms of what they are designing.

CHAIR: Does anyone want to add to that?

Prof. Walsh : This is a really important issue. It's worth pointing out that it's still a research problem. We don't understand this. Some of the reason why people are saying, 'We don't know how this algorithm works,' is that we don't know how algorithms work. We are still working out how to build algorithms that have transparency, that can explain their actions. This is especially true today, when we are increasingly seeing aspects of machine learning where the decisions are not just the product of what the designer, the computer programmer, made; they are also a product of the data it was trained on and the risks if you expose it to the wrong data—and understanding what might be biases that exist in the data. We will run into the problem that we will bake in societal biases, if we're not careful, by training it on historical data that has those biases, maybe sexist, ageist or racist in some way, which we don't want to perpetuate by putting it into a black box. These are issues that we still don't have complete answers to. They are ones that we should be very concerned about because they are deciding what jobs people have, what work people have.

I also want to raise one other word of caution about the immense network effects that are happening today. The companies that have the data, the platforms that have the data, are immensely powerful, and this is causing immense distortions. I'm going to be rude about companies like Uber because they are not returning the benefits to the people who are providing the work on that platform. There are two fundamental reasons why that happens. Why is Uber not sharing its benefits with the people actually providing the transportation? The first was that we needed some centralised mechanism to have a reputation system. No-one is going to get into a random car and no-one wants a random person to get into their car. Actually, that technical problem has now disappeared because we have things like blockchain which allow us to do that in a completely distributed fashion. The question is: why don't we have some sort of cooperative that does exactly what Uber does and shares the benefits with all the players? The reason is the distorted impact of venture capital. Uber doesn't try to make money. Uber loses money hand over fist. One dollar out of two dollars that goes into Uber is lost—every one dollar out of two dollars of every fare is lost money. They don't care about making money. So it's hard to compete against a business that is not trying to make money.

Mr Sharp : Can I jump in here. I just want to respond to the question of platform ownership, which relates to the question of algorithms and also the ownership of data. When the people that are creating value in these marketplaces have a stake in how those algorithms are developed and how that data is owned and managed, I think that is a form of empowerment in itself and can lead to those labour exchanges giving workers and other value creators much more of a say in how the algorithms and the data are put together and the outcomes they lead to. On the ride-sharing question and the question of cooperative ownership, we are starting to see examples like Green Taxi Co-op in Denver, Colorado provide something of a response to the challenges you talked about earlier in terms of Uber. That's a member-owned taxi cooperative, and 800 drivers have a stake in the business and are able to distribute profits back to themselves and have a say in how the company is run, and they have about a 40 per cent market share in that city. Obviously Uber is in hundreds of cities around the world. It is not necessarily about providing a replacement to Uber but just an alternative that can keep jobs and wealth creation in local communities through cooperative platform ownership.

Dr McNeill : Briefly, just to reinforce those points, I think what we're talking about here is issues of choices. When these decisions are made they are choices within organisations. When we look at that enterprise level they are choices that relate to capital when we're talking about shareholder owned corporations. When we look at cooperative models they are generally based on adherence to a particular target group or a particular place based community, so they are very much interested in transparency and having those kinds of decision-making things being very clear to people. One of the things we've found in the manufacturing research project is that, when we look at cooperatives like Norco up in northern New South Wales, who have been around since the 1880s and have survived all kinds of different ups and downs and changes in technology, the decisions they make through their cooperative structure and their adherence to the seven cooperative principles have allowed them to maintain that longevity and sustainability over time in ways that are very different to the alternative model within the same industry. In the dairy industry they had partnership with a competitor and they have ended up buying back that business into their business because the other shareholder owned firm couldn't compete and survive in that space. Looking at those cooperative models that have an adherence to those seven cooperative principles is a good way to start in terms of how you look at the decisions being made and the choices being made at that enterprise level, which then flow on and affect all of these issues that we're discussing.

Senator STEELE-JOHN: I have two questions; both are about cooperatives or cooperative-like models. I think I will go to Mr Sharp first. You said a couple of times during your statement that the idea of letting contractors have a share in digital platform companies would give them more of a say in the overall direction and therefore address some of the imbalance. Have you undertaken, or do you know of any comparative work between, say, the cooperative taxi company that you mentioned in Colorado and the rights and conditions of the workers within that system and a worker within a similar industry but under a traditional system?

Where I am driving to with that question is that we hear a lot about how it will give people more of a say, but I'm wondering where that more of a say lands in relation to what you would get, say, under a more traditional framework if you were represented by a union?

Mr Sharp : Sure. We're starting to see lots of different examples in the cooperative platform space. There is a catalogue of this, called The Internet of Ownership. It now has hundreds of examples of platform co-ops in existence around the world. What they provide from a governance point of view is a 'one worker, one vote' kind of thing so that the owners of the platform can actually have a stake in the decision-making of the enterprise itself. An example is in Canada, just to choose one: Stocksy United, which is an artist owned cooperative which sells stock photography online. It's a competitor to the likes of iStock, Getty Images and so on. On the governance question: as a member owner, the artists have a stake in how their company is run. From a prosperity and wealth creation point of view, the artists actually share the revenue—50 per cent of the licensing fee—with the platform itself. So there is a much higher degree of profit sharing that's going on.

From both a wealth creation point of view and from a governance point of view, these types of platforms that are owned cooperatively are providing the workers themselves with those benefits. Again, there are hundreds of these that are catalogued on The Internet of Ownership. Academics, like Juliet Schor in the United States, have written extensively about this. I can refer the committee to some peer reviewed literature on that afterwards, if you would like to have that on record. I can send it to you.

Senator STEELE-JOHN: That would be great. In terms of the contractual arrangements that are in place within these cooperatives, how do they interact with the social safety net and worker entitlements programs within the various countries where these things operate? Are people able to access things like annual leave? Or is the main benefit a greater wage share?

Mr Sharp : I might just defer to Professor Morgan. Bronwen, do you want to pick that up, possibly?

Prof. Morgan : I think this is a question where we need some sustained empirical research. There was a recent grant done by the Australian Research Council, to look specifically at employment conditions in the gig economy setting in Australia. That is not held by us; we've just put in a grant application along these lines, but not so much focused on the specific question that you raised. To pick up on one of the examples that Mr Sharp just mentioned, which is Stocksy, in that case the interaction is, in a way, solved by the same strategy that the large platform companies use with very different effect, which is to say that they're not employees; they're member-owners. Therefore, what they access, in terms of social benefits, is not an interaction with employment. If you were able to access a pension, what you declare is your personal and private situation with the tax authorities. You're employing yourself. The difference is that you really are employing yourself and you have a say over the infrastructure and the design environment, which Professor Tonkinwise was talking about, and how that operates. You have much more control over your conditions of employment. In the case of Stocksy, it really is global. They are aware of this question and see it as, in part, a challenge but one that they don’t address directly as an enterprise. Their members come from all different countries, so it's not possible. By doing research, it would be interesting to see how those individuals solve those problems in different country settings.

Senator STEELE-JOHN: Professor Walsh, you were talking about the way in which China's betting its future on AI. What does that look like on the ground, in terms of government settings and particularly looking at the curriculum they're pursuing? Is there a far greater emphasis on STEM or are they moving more towards arts education? What does that look like in reality?

Prof. Walsh : I can certainly speak to the higher education question, which is that there is a huge great emphasis on STEM, and they're now turning out three times the number of graduates in STEM than they were. They are now, I think, the largest STEM graduate provider on the planet. It's well exceeded that of the United States. So, yes, there's a huge great bet on STEM.

I don't want this committee, though, to think that all the jobs of the future are in STEM. There will be lots of jobs, in the future, in STEM—inventing the future. Equally, there'll be lots of jobs in artistic and artisan settings, where we'll appreciate things made by touch, by the human hand, and there'll be lots of people focused jobs, where emotional intelligence will be very important. We'll prefer our doctor to be a human; we'll prefer our psychologist to be a human; we'll prefer the sales person to be a human. STEM is really important. If we don't educate people with STEM and if computational thinking is not a fundamental part of the curriculum, it's going to be magic. And, if it's magic, you're going to be taken advantage of. You need a STEM-literate population so that they can take advantage of the benefits of technology. But we don't need lots of programmers. In fact, eventually, computers will program themselves. It's a pretty systematic logical process. We will work out how to get computers to program themselves.

Senator STEELE-JOHN: Do you think that, in some ways, there's a view of the focus on STEM as a kind of panacea when, really, it needs to be a baseline upon which we invest in things like arts education and social and emotional intelligence—

Prof. Walsh : Yes. Sometimes it's called STEAM—putting the A back in STEM. I think that's very important. There's an argument that we shouldn't be teaching calculus; we should be teaching computation in schools. It's as fundamental as that. It really is changing our world. The calculus and derivatives used to change the world when we were living in a much more mechanical world. We're living in a very computational world, so everyone needs to understand how those programs work and what's technically possible, but not everyone needs to be a programmer.

Senator STEELE-JOHN: But calculus is so useful in everyday life!

Prof. Walsh : You use far more discrete computation today. Of course, your car uses lots of derivatives and lots of calculus. For most of us, computation will fundamentally impinge on every aspect of our life. To be an active participant and citizen of society, I think you have a need to understand computation today.

CHAIR: You lost me at calculus!

Senator PATRICK: Professor Walsh, to segue from what you've just been chatting about, you've made the comment that you think computers will be programming themselves. My experience in working with a whole bunch of coders is that it's a very cognitive domain and involves the integration of algorithms that are generally done by PhDs and very, very smart people. Indeed, it's the very thing you were saying about data—trying something out on a piece of data, modifying the code, broadening the data that is then used until you get it right. How far off do you think that really is: the idea that a computer might code itself or start writing code?

Prof. Walsh : To us, in some sense, it's decades away; in some sense, it's actually here today. It's decades away before literally you'll just say to a program, 'Go and write a program to play an arcade game,' or 'Compute the optimal stock portfolio for me.' It may be decades away before you can just say some English and code will spit out at the other end. But machine learning is all about programs that learn. No-one knows how to program a machine today to recognise an image to drive down the road. The program taught itself how to do that, so machine learning is actually, in some sense, that dream of computers teaching themselves how to do a task. We don't know how to break that task down into code; we just say, 'Well, here's the basic framework of machine learning. Let's just train it on some examples,' and eventually it learns a program that does the thing we want it to do. That's very much how programming is changing, in some sense. The old school of programming—writing a database program or something—was certainly cognitively very demanding.

Senator PATRICK: Just stepping out of that, a lot of the submissions that were made had a common theme, and as I read through the submissions there was a sort of a socialist lean to the submissions that most of you had written. The next witness, the Australian Chamber of Commerce and Industry, is a capitalist, it appears. Being from a centralist party, I get both sides of the story. I'm not sure how far we got yesterday; I had to leave the chat about industry 4.1, but the German government had a plan that they laid out for 4.1. Are we simply lacking a plan about how to move forward with this from government that deals with both the socialist aspects of it and the capitalist aspects of it to get the right balance? I will throw that to the panel. I'm not sure who to ask the question of, but it seems to me that might be lacking.

Prof. Walsh : That certainly was the final recommendation in my submission: we need a government-wide plan. It doesn't just touch the future of work; it touches every aspect of government from education to how the military is going to fight in the future. Everything is going to be changed by these technologies.

Prof. Morgan : I would push back a bit against the socialist-versus-capitalist contrast, particularly through the cooperative lens. It is really a blend, and that's one of the things I said in our opening remarks. It bridges this polarisation, because it is an economic enterprise and it may have more of an emphasis on collective agency on the one hand and on pursuing non-shareholder profit maximisation goals directly on the other. But, if we take as important that those other considerations matter, addressing them through cooperative enterprise is actually much more on the capitalist side than addressing them through entirely state-down regulatory plans or fully nationalised business, for example. But I think we have heard a common theme about policy joining up across sectors, making sure that welfare and housing and income support and economic policy support each other. I think the idea of a whole-of-government plan is important, but with a very different take on it—that planning is not coming from the top; it's a whole-of-government look at what is needed from the bottom up—and that's what bridges these.

Senator PATRICK: I'm not sure if you're aware of the Senate inquiry into co-ops and mutuals that was run over the last—

Prof. Morgan : Yes, I did mention that earlier.

Senator PATRICK: The government, to their credit, have taken that and are working out ways to encourage co-ops. So that was clearly another theme that ran through all of the scenarios. I didn't mean 'socialist' in a negative sense; it was simply to say that all sides actually need to be properly considered.

Prof. Tonkinwise : Can I jump in and just say I would certainly agree that we need a plan, and we need a plan because a lot of the largest players in the platform economies in terms of the sharing economies have revealed that they definitely have nefarious plans to evade current regulation. I would just draw attention to Uber's use of Greyball in order to deliberately avoid regulators on the street, and I would just use that as an opportunity to also say that the people who were actually making that happen were designers. They were interaction designers deciding what, from the algorithms, was visible to whom. These aren't STEM people and they aren't humanities people; they're this weird blend in the middle, and they have enormous agency because of that mediating role. It's those interaction designers who are actually enabling the plans of these companies to evade regulation. So, I definitely think we need a plan in order to counter those kinds of initiatives.

CHAIR: If any of you would like to add anything further that you didn't get a chance to go into, feel free to send us something in writing. Senator Steele-John, for instance, was interested in getting a bit more on—

Senator STEELE-JOHN: The Greyball phenomenon.

CHAIR: Thanks very much to everyone for coming along today.