Thanks very much. Could I begin by acknowledging the traditional owners of the land on which we meet and pay respects to elders' past, present and emerging.
It's wonderful to be here in front of a room with so many illustrious people from an important sector in our economy. Can I just make a few very brief acknowledgements: firstly, can I acknowledge that we have many of the key regulators from the banking sector and the financial services sector more generally here, including APRA, and I think that John Lonsdale might be here or about to arrive, but also senior leaders from ASIC and AUSTRAC attending today. We have the head of the ABA, Simon Birmingham, here and as well many leaders from many of the key players in the sector - 2 of the big 4 banks' CEOs, but leaders from many organisations right across the sector. And can I acknowledge all of you. But, of course, just all of the people from not just the banks themselves but many of the other key players right across the banking ecosystem.
And can I, of course, acknowledge the AFR for organising today - James Chessell, I really should acknowledge Cosima Marriner, who is about to question me, just to make sure that my acknowledgement hopefully takes the edge off some of the questions. But, look, to acknowledge the role the AFR plays today but in other forums as well as in trying to bring together sectors and thought leaders to help us move forward with complex debates.
Can I just say that today I'm hoping to discuss some themes that I'm sure will be key themes discussed throughout the course of this forum, but I'm hoping to discuss it through a lens which doesn't replicate what people more expert than I am will discuss in terms of some of the applications of AI and other technology, and the productivity challenges at a very granular level. I'm going to try and talk about some of the challenges and opportunities that we're facing as a macro economist - and I should probably say 'lapsed macro economist', given that I haven't lectured or picked up a text book with intent for quite a while - but try and look at some of these issues as a macro economist firstly, but, secondly, as a public policy practitioner. And I'm hoping that by looking at it through those 2 lenses it dovetails with and complements some of the other analysis that we'll see today.
I was going to talk about 3 themes, and I was thinking about those 3 themes almost at concentric circles. The first theme was productivity in the broader economy and the banking sector. And that's the broadest circle. Secondly, AI in the broader economy and, again, in the banking sector, as a significant part of that productivity opportunity but not the only part of it. And, thirdly, to look at a case study, being scams, which I know is front of mind for many in this room and certainly for me.
When I think about productivity I think about the fact that potential gains of lifting productivity across the economy, even by a small amount, if sustained can generate huge gains in welfare over time. And I think about 2 very insightful people - one, Michelle Grattan, who I think late last year made the observation that all the galahs in the pet shop are now talking about productivity, which let's us know what we already knew in a sense, which is that productivity is the centre of the productivity debate at the national level but also in all of our organisations and firms.
But also, I go back to the Krugman phrase that productivity isn't everything, but in the long run it's almost everything. And that's his way of eloquently describing the fact that if you achieve very small gains in productivity and can sustain them, it really is the key part of what you can achieve in terms of welfare.
I would argue that in the services economy, productivity is both more difficult to define and to achieve than productivity in economies which are more based on agriculture and manufacturing. And I would say this for a couple of reasons. One is I think productivity is more difficult to achieve because I would argue that in a services economy, each sector tends to be more idiosyncratic in the productivity challenge and in the productivity opportunity. And that's an issue both for practitioners in those sectors but also for regulators.
I would also argue that productivity is more difficult to define in services because they are multifaceted, nuanced and, in many cases, debatable. And let's take banking. Now, I don't want to pose the question I'm about to pose in front of a room full of bankers to suggest that you haven't spent a whole career thinking about this question, but more I pose the question to indicate that I think it lies at the heart of the macroeconomic and the public policy question and is, therefore, worth explicitly posing. And that question is: what are the key outcomes that banking is seeking to achieve? I believe that it's only when you're very clear on the outcomes, particularly in a services context where they're not always obvious, that you can really grapple with the productivity question.
Now, there are too many outcomes for a brief contribution like mine to fully grapple with them, but I'll focus on 3: the first outcome I was going to touch on is the fact that banking helps individuals and families accumulate and spend assets over the life cycle. The savings book linked to a first saver's account that many of us encountered as children prompted so many of us to start that journey. A closely related outcome, of course, is to help individuals and families to borrow over the life cycle - buying one's first car or buying one's first home.
But how do we assess productivity in relation to this core outcome of banking? There are so many different characteristics. One would be the quality of services supporting accounts. One would be the individualisation of the experience that customers face. One would be the level and transparency of fees, the usability of apps, for some people it's the ability to access a physical branch for complex and sensitive issues. The speed and accuracy of credit assessments is clearly a key KPI. And then there's the respectful treatment of vulnerable customers such as the elderly or people from a CALD background. So I offer this as a non‑exhaustive list really just to show that even just in relation to this one outcome it's a very difficult outcome to measure let alone, of course, to provide.
A second core function that I think lies at the heart of what banking provides to our economy and our community is the intermediation of capital between savers and lenders, including to enable business formation and growth. But, again, how to assess the achievement of that outcome? One would be the efficient allocation of capital across the economy. And that's obviously an absolutely key theme in public policy. This has been a central theme at the various investor roundtables that the Treasurer has hosted over recent years, the last of which I co‑hosted one of the sessions of. And this was obviously one of the key themes of the three‑day economic reform roundtable, but not a straightforward KPI to measure either across the economy or even at an organisational level.
I would argue that another key aspect of achieving this outcome of capital allocation is the allocation across a range of purposes, including making sure that capital is accessible by different types of businesses, for example, that small business gets a go, or different risk profiles, that start‑ups gets access to capital.
And then finally, the third function I wanted to touch on is the role that banks play in our payments system. And this is absolutely central, obviously, to people's and businesses' day‑to‑day lives. Again, assessing the performance of organisations on this front is multifaceted and not simple. It includes the speed of transactions but increasingly it's not just lowering that speed but including appropriate levels of friction where there is risk of fraud or error. And this goes to the issue of scams that I'll touch on later. And people in this room day to day are dealing with this trade off where we're wanting to improve the speed of transactions while at the same time we're trying to build in protections.
There's obviously the cost of transactions, but then increasingly I think we're also having to deal with the issue of the fair distribution of allocating system‑wide costs across different cohorts. That's something which the RBA is looking at, which I know people in this room are well aware of. These aren't straightforward questions. And then, as John Lonsdale will pick up later, the management of systemic risks.
So, this is not an exhaustive list either of the core functions of banks or of the ways in which we might assess each of those functions, but I'm hoping that even in giving such a basic outline it demonstrates how thinking about productivity in relation to services is very difficult - but important.
And I frame this because I think it is important to go back to basics occasionally. I know that you do this as organisations frequently, but it's also important to go back to basics as public policymakers because we're looking at a lot of those challenges through the lens of public policy, and it's also important that we have a clear understanding of issues like outcomes and productivity when we're designing regulation.
Finally, on the issue of productivity, I want to link that to question of regulation. Much regulation can be productivity‑enhancing, for example, by building trust in markets or in transparently managing systemic risks. And, again, I acknowledge APRA's presence. Regulation can also help to achieve outcomes like consumer protection that competitive markets often won't achieve well on their own, particularly where there are low engagement or sticky customers.
But governments need to constantly think about the way in which we regulate. Are we achieving the desired outcomes of regulation without unnecessary red tape? Are we changing regulatory settings too frequently? Is regulation well designed for changing circumstances? And one that I think is key is that where there is innovation that is not well understood, are we managing risk in such a way that we are protecting consumers from emerging risks but not killing off the innovation itself? And this brings into play issues like sandboxes, and are we using sandboxes in the right way? And this is an issue that is currently under review by the government. For me, this is a critical issue of the interplay between regulation and innovation and productivity.
And finally, are we suitably aligning interventions by regulators, including data requests, for example, across agencies and across layers of government? And that's something which I've been working with many people in this room on.
So, I do think it's worth posing some of those foundational questions. And, as I say, I know that you do that often as organisations, but I think it's worth doing it as regulators and public policymakers also because it helps us to think about what we're trying to achieve and what this word 'productivity' means in a particular sectoral context.
So, the second theme I wanted to touch on was AI and all the related issues, such as digitisation and so on, again both from the perspective of the economy, but also the banking sector. Now, there's been much discussion of AI across the economy and much discussion of the potential to transform and to lift productivity growth rates. And, of course, we know that AI investment is already very significant in the banking sector and is having a transformative impact.
Much attention is focused on what is the trajectory of AI across a number of dimensions, including its impact on productivity, including its impact on labour markets, including its impact on public policy questions such as privacy or the creative arts. Now, I don't want to stand here today in front of a room of banking experts and talk to you about how AI is going to play out in banking. None of us really know that, but you would know more than me about that specific question. But, again, what I want to think about is from a broader perspective, how do some of the macroeconomic lessons of the past suggest we might think about AI and its trajectory across the economy.
Now, there aren't many iron laws in macroeconomics, particularly when trying to forecast something as complex and dynamic as AI. But there are some broader guideposts that I believe can be useful. There are 2 broad macroeconomic empirical trends that I think - well, they're both somewhat counterintuitive, but I think they're also potentially instructive. The first is the if you look at the GDP share going to labour and capital over the last hundred to 150 years in Australia but, indeed, in almost all advanced economies, it's been roughly two‑thirds, one‑third. And this has been relatively stable for a long period of time across most macro economies.
Now, I wouldn't claim that most macroeconomists have a clear sense as to why that's the case, but we have observed that it is broadly true. And most macroeconomists - macroeconomic models assume that to be the case and then build an economic model around that. But that economic regularity, I think, is an interesting starting point.
The second is that, again, across most advanced economies overall labour force participation has trended up over the last century, not in a linear or constant way, but if you look at it over the very long term from the early part of the 20th century to today, the trend in most advanced economies has been generally up.
Now, what is somewhat surprising, at least for me, is that these 2 regularities have held true even in the face of many significant economy‑wide disruptions. They've held true in the face of the mechanisation of agriculture in the early part of the 20th century. I think it's worth remembering that in many advanced economies agriculture still late in the 19th century constituted 30, 40, 50 per cent of the labour force. Now, that mechanisation of labour in the agriculture sector didn't happen very quickly, but it was a sustained, transformative impact on the economy. It held true in the face of the electrification of advanced economies in the 1920s and 30s and beyond. The automation of much of manufacturing, the IT and PC and telecommunications revolutions of the 80s and 90s and beyond and the incredibly transformative global interconnectedness arising from the internet in the early 21st centuries.
And I would just make the observation that each of these transformations occurred at different speeds, but each of them involved whole‑of‑economy transformations, and none of these materially turned the dial on these 2 higher‑level macroeconomic trends, which is that the majority of GDP was returned to labour and, secondly, that labour force participation in a non‑linear way has continued to broadly track up.
Now, what do these 2 empirical observations suggest for AI? Well, firstly, nothing for certain. I just want to make that point clear. But, while it's true that I am very confident that AI offers a great deal of productivity upside, I would also suggest that in relation to some of the material risks that people are observing, these 2 broader macroeconomic trends suggests that people with a bit of regulation and perhaps with a bit of prodding, we might imagine that we can see a plausible benign outcome on at least some fronts from AI, even if it is transformative.
So, first, if labour force participation has risen, or at least not fallen, in the face of the waves of automation and mechanisation and technological transformation that I just described, it's possible that with AI we will see something similar. We could see the creation of large amounts of jobs involving empathy, involving human connectedness. And we've all read different articles by experts outlining the possibility of that upside. I would simply say that when it comes to AI and the labour force, there are real risks, and we will see job losses in particular firms and particular sectors. But let's not forget the potential for upside, as we've seen in the past.
Having said that, we're going to need a lot of effort to help individuals transfer into a lot of those emerging areas. But let's not forget that our economy and our society has proved to be incredibly adept at creating new professions and new skills in the past.
Second, while there has been much discussion of economies of scale, of global and local winner‑takes‑all and network externalities of many new technologies, it is plausible that the lion's share, or the majority, of GDP will continue to reward labour even after AI's been introduced. Again, I'm not saying that's for certain; I'm simply saying there is something in the economy which has withstood earlier round of disruption, and it's plausible that that will continue, possibly linked to the earlier point that I made, which is that we continue to see new types of jobs created.
I should add a quick caveat here - that since the 1970s, the late 1970s, we have seen labour's share of income across the OECD fall from around 66 per cent to 60 per cent. Now, we're not quite sure exactly why that's happening, but it may be that AI accelerates that trend. So, there are some structural - there is some structural inertia, but there is also a multi‑decade trend already underway which might reflect skills bias, technological change or the growing capital intensity of certain sectors.
Now, even if these somewhat benign macro scenarios occur, I do think it's worth noting that there are some possible risks. Firstly, I do think that productivity will be positively impacted, but I also think that it's going to continue to be difficult to measure outcomes and KPIs in the services sector.
Secondly, I think that we could well see stark distributional outcomes as some form of labour and capital become more productive and some relatively less so. Again, this has been a feature of earlier transformations. I'm confident that it will be a feature of this one.
We could see some people having to retrain more often for jobs. The upside of this could be people having more diverse and more risk‑protected careers and more interesting careers. But what it will also require is that people be supported through those retraining needs, and this is where government comes in. We're already seeing a whole range of efforts by government to support this through free TAFE, through apprenticeships, through interventions in the university sector. But I think this is only going to increase as a societal need.
We could see more individualised service as a result of AI and digitisation, and greater data granularity. This will, on the whole, I think, be a boon for consumers. It will lead to more individualised service that caters to individuals 'preferences, and risks and needs. But there will be some circumstances where there might be downsides. For example, could more accurate credit assessments see some people excluded from loan markets? And in other contexts of financial services, we're seeing situations where more granular data is creating a risk where some individuals might be priced out of insurance, for example. So, this individualisation, on the whole, is going to be a good thing, but it could create some potential negatives.
And then I would say there are some areas where AI and technology may be a zero‑sum game. For example, in the area of scams - which I'll touch on briefing in finishing. Scams is an area where we see massive investment in AI by the perpetrators, but also a massive investment by banks and regulators. Now, if the end result is that we protect people and most of them don't see a scam eventuate, from a societal perspective, it may be that the net result of all this investment is not much gained but much‑needed protection. It's an interesting example of where investment won't necessarily be reflected in welfare gains in the same way that it would if it was an investment in a service per se.
Now, I want to just simply finish my observations on AI by saying that even as AI transforms many things, I don't think it's going to change the fact that banking will remain a business centred on people. I believe that banking is an industry where its employees are its most valuable asset. Even as an industry that invests in computers, in IT, in buildings, it is a business where its people are its biggest asset.
And, secondly, of course, it's a business that needs to take into account the humanity of its customers, their limited bandwidth for complexity, their individual needs and their occasional vulnerability. AI will involve technological innovation and capital concentration, but people will remain at the centre of our national productivity puzzle, and will remain central to the business of banking.
Now, I just want to finish on scams as a case study in the sense of AI - and so the third ring down in the concentric circles. The context of scams, as you all know, is that losses were rising sharply, and so intervention was needed. The government under my predecessor took the step of adopting an ecosystem approach, which I believe was the right approach, and it's world‑leading and being looked at by many other countries. But it is a complex approach.
We've already seen the SPF passed in early 2025. We have the NASC in operation, we have fusion cells in operation. Much is already happening that's positive, and we've seen actually a drop‑off in aggregate losses, even though that might be plateauing in some areas. The next steps are to designate sectors, including banking, and telco and digital platforms; it is to devise an IDR and EDR framework, which is going to be very complex, and it might involve the simultaneous and contested attribution of guilt across multiple parties - not guilt, a liability; and it will also involve complex information‑sharing.
But in the meantime, there's lots of good things that are happening, and I'll finish by looking at an elderly couple that came into my electorate office last Friday. And they're from a non‑English‑speaking background, from the Vietnamese community in my area. A third of my electorate is Vietnamese background. And they had seen a very credible AI image or video, rather, from a high‑profile politician saying they should invest in a particular thing, which was clearly not valid.
Now, after having looked at that video, after having done a bit of due diligence, they decided to go ahead. They reached out to their bank to undertake this transaction, and the bank then reached back to them and said, 'We really strongly recommend you don't undertake this transaction.' And they got angry. And, in particular, the woman got angry and then tried to get the husband to bypass this recommendation by seeking a different way to undertake the transaction. But, again, there was pushback.
Now, in the end, the transaction didn't occur. They then came into my electorate office, and I think they by then had strong suspicions, and we sat down with them and looked at the video and confirmed that it was fake. And we all had a bit of a laugh about it - about the fact that they really got quite annoyed about the fact that the bank had reached out to them and told them not to undertake the transaction. But on reflection, they were very glad that happened.
And for me, it was a very important reflection on, firstly, the fact that this is a human story ultimately. It's about technology, both from the perpetrators, but also the banks and the regulators and how they intersect in the lives of people who are not familiar often with the financial services system. So this was a good outcome, and it reflects the fact that there's already progress being made. There's lots more progress to be achieved over the course of the year. But this was a good sign that together through government, through regulators, through what the banks and other sectors are already doing, there is progress.
And when we look at that couple, I think it says a couple of things: one is, if we're thinking about what the outcome we're desiring here is, it's protection. And that was achieved there. But even that is a multifaceted and complex thing. How do you identify transactions, and which transactions get flagged? How do you reach out? And then the humanity of how do you reach out in a way that's sensitive, that is persuasive, that is appropriate, and how do you do that in a way where people who have valid transactions can undertake them?
Secondly, it reinforced to me that this is a really important area, but there are zero‑sum aspects to this where we're seeing massive amounts of resources as a society deployed on all sides of this transaction. Now, if the outcome is people are protected, that's a good thing. But it is interesting as a macroeconomist to think about how do we think about this from a broader welfare perspective. And some of the great minds of macroeconomics over the last half‑century have thought about how you think about such situations in terms of GDP.
And the final thing - the third aspect of that interaction and the thing I'll finish on is to say it reinforced to me that all of these public policy complexities that we talk about are ultimately about people. On this occasion, it worked. But it was a complex thing because the bank was having to deal with a complex situation and a high‑risk transaction but also dealing with individuals who weren't familiar with either the IT or the banking system. And we need to make sure that we continue to have the person at the centre of what we do, both you as banks, regulators and me and my colleagues as public policymakers, to ensure that we continue to get the best outcomes for individuals and society. Thank you.
Cosima Marriner:
Thanks, Minister, and thanks for joining us here this morning, nice and early. Your speech was very interesting on a lot of fronts. I might just pick up on your last point about scams and your very compelling anecdote about some of your constituents. Now, AI, we're reporting today in the Financial Review that there's concerns that the $1 billion mortgage fraud at CBA has spread to other banks. Should banks be doing more to prevent this kind of fraud, and particularly when it's being turbocharged by AI?
Daniel Mulino:
Well, I guess my answer to that question is I probably suspect that the answer to that depends on which sector we're talking about. I think when it comes to scams, my sense is that a lot has been invested and that Australia is world‑leading. And so I think there's more to be done. But my sense is we are actually pushing the envelope in that area. And when I've gone to visit banks and seen the teams they have - hundreds of people in teams - reaching out to people, flagging dubious transactions, again, more to be done, but that seems to be an area where we're leading.
When it comes to AMLCTF, I suspect probably we're not necessarily right at the cutting edge there. And I remember back when I was in opposition, Kristina Keneally as the shadow home affairs minister gave me a bit of scope to look into AMLCTF, and the tranche 2 legislation was front of mind - bringing real estate agents, and accountants and lawyers into the AML regime.
Now, we were one of the lagging jurisdictions, according to [indistinct] back then. Now, that has since been remedied, but what it says to me is - again, I comment on this without being the minister or an expert - but I suspect in that area, if what we're seeing turns out to be systemic, it may be that that's an area where more is needed and where we're not leading. So I suspect, like any big complex organisation, it probably depends where you look.
Marriner:
So you're flagging that there may be more legislation around and anti‑money laundering provisions?
Mulino:
No. I mean, look, my sense is that the legislation that went through last term with Mark Dreyfus as AG, that that fulfilled the obligation we had. I think if you go back 5 years to when I was in opposition - and I'm speaking on this - the then government had made a number of commitments to bring in that tranche 2 legislation. And, look, I'll acknowledge, there were some tricky aspects in that you're dealing with a bunch of stakeholders, some of whom didn't want to particularly want to be brought within the regime. There was a lot of commentary in the AFR at the time, I remember, with certain real estate transactions under the microscope. So my sense is that that legislation has now remedied that issue, that [indistinct] was very concerned about. And, again, I don't want to get too much into another portfolio's details, but my sense is that we now - to step through, I understand that AUSTRAC and ASIC and others are looking in detail at this. So until they have a close look, I wouldn't want to make too much observation. But this is clearly an area where organisations are having to build their capability.
Marriner:
Now, you give us a bit of an historic macroeconomic perspective on, you know, labour and capital and how you see AI sort of fitting into that historical suite. And you were in the main relatively positive. But you did say that there would be a bit of reshaping of the workforce. Now, we're obviously seeing some, you know, big job cuts flowing through, particularly in the tech sector. We've had Atlassian, we've had WiseTech, you know, and they're talking about AI being - part being the reason for this. There's a lot of talk about we may get to a stage where governments have to provide a universal basic income because there aren't enough jobs for people to do to earn money. Have you given that much thought?
Mulino:
So, look, I'm going to misquote an essay here, possibly, which is terrible, but I remember there was a famous essay back from the 30s that I think Keynes wrote, where he basically said if you fast forward 50 or 80 years, you're just going to have a distribution issue and not an issue of people working. But even if it wasn't - even if I'm misattributing, there are a lot of economists who have that thought.
But then, even more recently, in between me being in state parliament and federal parliament, I did a bit of consulting work on the automation challenge. And there were a whole lot of consultants and public policymakers who were basically looking at occupations right across the economy and trying to identify which occupations were going to be automated out of existence or which occupations where there were maybe some complement or substitute dimensions, which occupations would be half automated. And I remember all sorts of different firms had different approaches. But ultimately, a lot of them were trying to come down to some kind of number of what proportion of labour is about to be automated.
Now, this was pre‑AI, and this was all sorts of things like, you know, spreadsheets, big data, all the kinds of things that were going on in white‑collar work before AI had taken the next qualitative step. And I would simply say there are a lot of aspects of that debate that are different but not totally dissimilar to what we're going through now. And I must say, I always felt in that debate in, you know, the 90s, early 2000s, that people who imagined that a third of the economy would just disappear because of, you know, automation, were perhaps underestimating our capacity to adjust what we do, create jobs elsewhere. And I think that's being born out. I mean, there are all sorts of sectors in the economy which have involved human interaction - parts of the care economy and others - which have grown substantially.
Now, we could have a separate debate around the inherent productivity of those sectors, and that's another debate which is prompted. But there's no doubt that the economy has, for a long time, had quite a capacity for generating new things for us to do. So I would simply say I wouldn't underestimate some of the challenges we're facing, but I would also say that if we have an ambitious perspective when it comes to helping people retrain, if we have an ambitious perspective when it comes to helping new businesses form, I would think a relatively benign outcome in terms of labour force participation is possible.
And, sorry, just on the UBI issue, my sense is that rather than a UBI, which is going to take our existing I think quite effective welfare system and then just add massive costs to it, I would rather figure out are there gaps in our current welfare system where we need to expand it or maybe provide a slightly better or better designed safety net. I feel that's going to be better bang for our policy and taxpayer dollar than something universal, is my gut instinct.
Marriner:
But you think there may be a need for an expanded safety net if AI does sort of transform the labour force?
Mulino:
Well, I think even before AI came on the scene - and, again, I don't have figures to hand - but the average number of careers was creeping up. Even before AI, when I read a lot of public policy work around the automation challenge, there was a sense that, as a society and as public policymakers, we were going to need to think a lot more about mid‑career retraining. And so if you look at, for example, the lifetime accounts in Singapore and Denmark, a lot of those kinds of public policy developments preceded AI.
So my sense is that a lot of the things that we're going to need to do are already in train. I think free TAFE has been really important. And one of the things that amazes me about what's going on in free TAFE in my area -at the dual provider at VU - is that there are more people at the moment who are taking - with a bachelor's degree who are getting a TAFE course so as to go into a career that excites them where there is ready demand for jobs than there are people who are looking to upgrade a TAFE qualification to a bachelor's.
Now, I'm not making any kind of aspersions or comments more generally about the economy, but it's simply to say that this issue of mid‑career training is a huge opportunity. And I think it was there before AI. It may be that AI means we have to invest more in that. I think that's a real opportunity.
Marriner:
So let's take a step out. We've got the Middle East war continuing. We have the US Energy Secretary this morning say that he says the war ending in weeks. Do you share his confidence?
Mulino:
Yeah, well, I think in question time last week, one of the ministers was asked when is the war going to end, and I think he politely declined to give a precise figure. So I think what I would say is that - and this will sound like an obvious statement, but I simply figure I think it's worth putting out there as a starting point - is that pretty much anybody who's been undertaking macroeconomic analysis or macroeconomic modelling would say that the key determinants of the risks to our economy, to inflation or to other aspects of the macroeconomy are the duration and the scope of the war.
So I think this has been something which has been commented on by a number of macroeconomists and modellers over recent days. And this is the challenge. I don't think it's easy to pin down how long these kinds of conflicts will take once they get going. There are, you know, obviously going to be some impetuses for a number of parties to try to look for resolution if the economic costs rise. But all I would say is that, clearly, we're hoping for an earlier resolution than a later one.
Marriner:
Now, the Treasurer said yesterday that the war makes reform even more urgent, and he's framing the Budget as a reform budget with, you know, productivity measures in it. There's been a lot of, kind of running things up the flagpole, so it would be good to get a little bit of more definitive answers on some of them from you. Capital gains tax discount, are you going to scale it back, and will it be retrospective?
Mulino:
Yeah, I mean, you won't be surprised that I'm not going to announce budget measures today. But, look, what I would say is that I think the Treasurer is spot on - and you won't be surprised to hear me say this - but I think the Treasurer is spot on in that we need to take the opportunity for ambition and reform.
I would say that the three‑day reform roundtable last year was an incredibly useful forum in a number of ways. One was that it led to practical pro‑productivity reforms. So I don't think the EPBC Act would have passed late last year had it not been for the momentum coming out of that. It led to the freezing of the NCC and a whole bunch of other important things.
But the second thing I think it did was to help us across a very broad constituency and range of participants, frame how we should think about reform, including tax reform. And I think that some of the touchpoints that came out of that for me - and I think these are things which the Treasurer has touched on - include the need for us to focus on intergenerational fairness. And that's something which I think has featured in a lot of the discussions. I think the other thing is that we need to think about the simplicity and efficiency of our tax system. And again, I'd go even beyond tax, and I touched on in the speech on the fact that the Council of Financial Regulators has a broader project when it comes to the design of regulation more generally.
And the third is I think, we need to think about the role of taxation in investment and business formation. So I think what I would say is that those are 3 themes or touchpoints that I think will inform a lot of the conversations over the next few weeks.
Marriner:
Can I just pick up on your last one about business investment? Our splash on the front page of the Financial Review today is about Google threatening to abandon plans for a $20 billion investment in data centres in Australia because it's concerned about the tax treatment that that might expose it to. Do the tax laws need refreshing for AI and data centres?
Mulino:
So, look, that's a live discussion and a complex one. And I don't want to get into the details of that one. But what I would say on data centres is I was recently at the opening of a CDC data centre in my electorate - you know, multi‑billion dollar globally cutting‑edge technology in terms of the cooling. Again, I don't want to sound like a cliché with these things, but it's a very data and capital‑intensive industry, but it was refreshing to me to see hundreds of people in the audience who work at CDC. So a lot of these industries are capital and technology intensive, but there is that people dimension.
But, look, one thing that I gathered from the discussion with the people there, but also others in that sector, is that if you look at the data centre industry globally, by some accounts, we are the second most attractive jurisdiction. I've heard from some sources that for a whole range of reasons, including our stability in terms of the regulatory and other settings, in terms of access to capital, a whole range of reasons, we are a very attractive destination for data centre investment. And I think that reflects Australia on a whole range of fronts.
So the issue that was on the front page of the Fin is an important one, and that will be worked through. But I would stress that if you look at the amount of investment happening and the amount of investment in the pipeline, Australia is right up in the top echelon of jurisdictions at the moment. And that's both a kind of qualitative assessment from a whole bunch of experts, but it's I think being born out in the numbers.
Marriner:
The flip of that is the Australian Banking Association - and Simon Birmingham will probably talk about this in a little bit - is, you know, there's an argument, particularly with the banking sector, that the big tech companies get an easy ride on regulation and tax. How much sympathy do you have for that argument?
Mulino:
Yeah, I mean, look, I would say - and this doesn't go specifically to the tax question, but on the tax question, I would say Australia has done a lot to strengthen multinational taxation and a lot of the initiatives going through the OECD. But, look, happy to kind of hear people's thoughts on that agenda more broadly.
But I think when you look at the big digital platforms more broadly, I think you'd be hard pressed to mount an argument that Australia isn't at the forefront of regulating big tech when it comes to a range of their interactions with society and impacts on society. So, you know, the under‑16 ban is one example. In my own portfolio, I'm working on the news bargaining incentive, which is the relationship between some parts of big tech and, you know, public interest journalism. It's a very complex area, and what we're doing is quite different to what other jurisdictions have done, where they've [indistinct] digital services taxes. We're trying to do this in a much more, I would argue, nuanced and targeted way, where it's not about raising revenue; it's about encouraging parties to come together.
So I would argue that on a range of fronts we are the jurisdiction that is leaning forward. I would say that, you know, we're a jurisdiction that wants the benefits of the digital platforms and that sector. There's clearly a lot of upside. But, you know, we've already lent in in a lot of ways that is world‑leading. And, again, going back to the under‑16 ban, there are a lot of countries now who are looking to implement something similar. So the tax point, I acknowledge, is ongoing.
But I think more broadly - and then the only other point I would make is - and again, the relevance to this room is when it comes to the payments system, we've seen a number of laws passed just in this last 11 months or so, which will give the RBA significantly enhanced powers to bring different things within their remit when it comes to the payments system. So I think that's a useful thing as well because the payments system is an area of such innovation, I think making sure that the regulatory system is fit for purpose is critical.
Marriner:
Minister, we could keep talking, but we're definitely out of time. Thank you very much for coming today. We really appreciate it.
Mulino:
Thank you very much.
