THE FUTURE OF WELFARE IN AUSTRALIA

Economist and Labor MP Daniel Mulino for Fraser is also the author of Safety Net – The Future of Welfare in Australia. In his book, Daniel Mulino explains how the Australian welfare state was created – and what we need to do to protect and extend it. The welfare state is one of the crowning achievements of the twentieth century, giving citizens access to healthcare, pensions, disability and unemployment benefits. This unprecedented expansion of the state was a product of the post-war period of the late 1940s, when governments ramped up investment in this grand safety net. By the 1970s, half of all government spending went towards social-welfare programs, but today the welfare state stands at a crossroads, beset both by political opposition and funding pressures as the population ages. On Wednesday 24 August, Daniel Mulino addressed The Sydney Institute to expand on some of the issues facing governments in the 2020s

THE FUTURE OF WELFARE IN AUSTRALIA

DANIEL MULINO

Thanks very much, Gerard, and thanks also to Anne, Jacquelynne and The Sydney Institute for giving me the opportunity to say a few words here today. Can I acknowledge the traditional owners of the land on which we meet, and pay my respect to elders past, present and emerging.

I’ll start with the concept of risk and make the observation that between 2009 and 2012, governments in Australia spent 50 times more on cleaning up after natural disasters than preventing or preparing for them. During this period, the Productivity Commission has observed that federal, state and local governments spent around $11 billion on post-disaster relief and just over $225 million on mitigation and resilience programs. As for Covid, the failure to sufficiently prepare for the risk of a global pandemic was astronomical in both financial terms and the loss of life.

In 2005, Michael Osterholm wrote in the New England Journal of Medicine of the need to prepare for the next pandemic. In my words, pandemics are white swan events not black swan events, borrowing from Nasim Taylor; we know there are going to be more pandemics. In Osterholm’s words: otherwise, the “loss of human life will be devastating and the cost of a world economy in shambles for several years can only be imagined.” Prescient words.

The imbalance in what we spent in the lead-up to Covid on medical stockpiles and preparedness, and what we had to spend afterwards compared to our expenditure after the pandemic hit is almost incomprehensible. The federal government spent over $350 billion directly, state governments spent a great deal of money and, of course, we had massive indirect costs, both in terms of the economic slowdown and the human cost. The imbalance between what we spent on preparation and what we had to spend during and after Covid, I believe, would be far greater than 50:1 if we were to do a rigorous analysis. We clearly need to invest more in preparation for future natural disasters and pandemics.

But this raises what is for me the even more significant, challenging and difficult question and, I believe a more important question – both for society and the federal budget: how can our welfare state better manage risk for individuals? Risk management has always been central to the welfare state but more recently has not had the focus it should. The welfare state in fact emerged in the late nineteenth century precisely as a result of regulatory reforms to existing insurance schemes.

We now stand at a fork in the road, facing a $1 trillion debt following unprecedented levels of government spending to protect the livelihood of citizens and keep them healthy, sustained low productivity growth and rising cost pressures across our social welfare institutions. It is time to adjust how we deliver welfare to ensure that the welfare state continues to exist as a meaningful social safety net.

Welfare is the foundation of all our lives. Everyone benefits from a strong welfare state, in our day-to-day lives and in national emergencies. The pandemic exposed just how much we all – rich and poor – rely on key public social institutions. During covid, all relied on vaccines, testing, a strong public hospital system, well-trained doctors and nurses and allied professionals. Countless lives were saved.

A modern, productive economy also relies on a strong welfare state; the sexy, high-tech industries that feature so prominently in our consciousness need healthy and educated workers that relied on the welfare state’s support through their lives. These companies rely on social insurance just as much as the public and not-for-profit institutions that are dedicated to alleviating poverty.

Unfortunately, when most people think of the welfare state, they think first of the unemployment benefit. Yet this payment (at less than $20 billion a year in a typical pre-COVID federal budget) is a fraction of what the government spends on the welfare of all its citizens and is not an accurate reflection of the flows of resources of the welfare state on the whole. The two biggest areas of government spending by far are health and the retirement income system – with both representing many multiples of the much more high-profile unemployment support. With an ageing population, enormous pressure will be put on both of these two programs. Of course, those payments are important, and we need to design them well. But I feel, at times, the public policy debate, and especially the economic literature, is skewed away from where we actually devote our resources and where we have the biggest impact on people’s lives.

I believe the welfare state has three key intellectual pillars – the universal provision of services; the redistribution of resources; and a focus on risk management. All are important, and often they’re aligned, but they’re not always aligned. A lot of what I explore in the book is where we can achieve better results, by putting a greater emphasis on risk management and insurance.  Yet as the above examples illustrate, governments tend to adopt a reactive approach to the many risks that society and individuals face.

It is instructive to think of the welfare state as a vast social insurance scheme. We all pool our resources to help those who through no fault of their own need assistance after fate has dealt its cards. Nobel prize winning economist Paul Krugman famously described the post-WWII US government as an insurance scheme that employs a standing army.

Framing the welfare state in this way is also a useful way to think about future reform. First, as noted earlier, it is a more accurate reflection of the key flow of resources in the welfare state than is the undue focus on the welfare state’s more controversial elements.

Second, if we think of the welfare state as largely or wholly redistributive, from the rich to the poor, it can create a “them and us” mindset – a class of people who give and a class of people who receive – “lifters” and “leaners” to use Joe Hockey’s words. The “makers” vs the “takers” in the US, and the “strivers” and the “skivers” in the UK. But for the vast bulk of our social insurance institutions, that’s not how things actually work.

We all benefit from this vast insurance scheme, particularly in the areas of health and retirement income support, which comprise the biggest spend of taxpayers’ dollars. Not only that, but in a range of longitudinal studies, both here with HILDA, in the United States with the PSID, and right across the world.

The Productivity Commission undertook a report into this not long ago. Many of us spend our lives moving up and down the income and wealth deciles. Unfortunately, there is a class of people who are stuck at the bottom and, for them, insurance is the right approach. The world is not about “lifters” and “leaners”, or “makers” and “takers”; the vast majority of us through our life, through the life cycle, through life events, find ourselves taking different benefits at various stages of our life cycle. I believe insurance is the best way to think of that.

When we think about unemployment, for example, so much of the academic literature ends up trying to model the mindset of an unemployed person as an equation, where the length of benefits, and the amount of benefits are two variables who dominate their thinking at the margin. And if we increase the benefits a little, they’ll be more inclined to stay unemployed longer. If we increase the duration of unemployment, they’ll be inclined to stay unemployed longer. Moral hazard matters in the real world, and incentives matter. As an economist, I don’t want to suggest that those models have nothing interesting to say. But I believe, particularly when it comes to our long term unemployed, the people who come into my office whose lives are stuck on a treadmill, the people who desperately want to work but can’t see a way forward.

Those people need investment. Those people need a different approach. Whether the unemployment benefit was 5 per cent higher or 5 per cent lower, whether we slightly increased the benefit or punished them a bit in order to make them work harder, whether we make them apply for 20 jobs a month – I believe that investment is a better approach; and investment flows from an insurance mindset. I could take you through any number of academic models – each capture something true about the unemployment system. But I believe, more often than not, those models, where people are making decisions at fine margins, about benefit levels or about the matching in markets; how picky they are about jobs depending on how good a match they think they can get if they wait longer, I don’t believe those models work well for our most disadvantaged people.

This can matter, both regarding our terms of reference but also in relation to how we conceive of and evaluation programs. Take unemployment:

• When one thinks about it as a transfer payment, so much of the emphasis of analysis relates to potential moral hazard. How should we design a program to reduce incentives to shirk? How should we design a program to reduce people being too picky? What will the impact of time limits be?
• In contrast, if the frame is the long-term investment in a person to restore them to their previous position, one thinks different. Joining up different services so that labour market programs are about not just short-term income support (important!) but also retraining.
• Consider a focus on how to structure short-term payments vs effort put into how to make lifetime accounts work effectively.

From the earliest days of welfare, it was about insurance. Even before modern governments, societies had to manage risks. There was food sharing in agricultural societies, and historians have found quite sophisticated ways in which people shared food, which shouldn’t be surprising, given it was a matter of life or death. In Greek and Roman times, occupational organisations, much like the guilds, looked like modern unemployment insurance. Private insurance emerged in the Middle Ages, once a bunch of geniuses in Europe found out how to manage the ins and outs of probability. Then after John Graunt and Edmond Halley started collecting data, it was that combination of probability and data which allowed private insurance markets to evolve in life, fire and other areas.

It was that patchwork of somewhat ad hoc arrangements that the State then started to expand dramatically in the 1880s. Key bills introduced by Otto Bismarck’s government in the 1880s all included the word insurance. Insurance for health; insurance for income in retirement; insurance for workplace accidents; and insurance for disability. We think we’re reinventing everything when we expand a social program, but if you look back at those days, there was quite an expansive vision. The difference between what Bismarck did and what we do today is in how a lot of those arrangements weren’t heavily subsidised, so they weren’t as much a cost on the State. But the vision, nonetheless, was very much around risk management.

So too was Franklin D Roosevelt’s core Second New Deal reform in the midst of the Great Depression – the “Old Age, Survivors and Disability Insurance scheme”, which remains to this day the United States’ most effective anti-poverty measure. He talked about the fact that existing arrangements were underfunded, inefficient and sporadic. I looked at that sentence and went back to the Productivity Commission report into the NDIS, and it was almost the same sentence with slightly different adjectives. For me, the welfare state has been about taking existing risk management arrangements, broadening them and making them more coherent and comprehensive. In Australia, we were early adopters; we mandated participation in a number of schemes around the turn of the century. State governments implemented old age pension schemes. In 1908, we had Deakin introduce the old age pension, which was legislated by Fisher.

There were great debates in Australia in the 1920s and 1930s around national insurance. Ince and Kinnear. The great debate was over how to fund national insurance; the non-Labor side wanted it to be risk rated contributions, and the Labor Party wanted it to be funded by progressive taxes. Going back to that time, we were ahead of the Beveridge Report in the UK, for example, and a lot of other developments in Western Europe. Then, of course, there was Beveridge, who was ground-breaking and influential globally.

Many of the social welfare institutions with the most impact are structured around the core regulatory interventions that are designed to overcome limitations of private insurance markets: mandating participation; subsidising participation; or limiting when and how much premiums can differ across individuals. These go to the heart of what our welfare state is about.

Historical studies of speeches by US legislators advocating the expansion of social insurance schemes found that they referred to core economic risk management principles such as adverse selection and moral hazard, even if they wisely eschewed the economics jargon.

Now is a good time to step back and reimagine the welfare state. As I mentioned earlier, we have a trillion dollar debt, we have a welfare state that is a very significant proportion of GDP, we have cost pressures in the very largest schemes, in both our health system and our pension retirement system, because of an aging society and cost pressures across both of those schemes. Despite efforts by all sides of politics, the productivity challenge is not an easy one. We are in the midst of a sustained period of low productivity growth. The key rationale for the welfare state – to protect us all when we are vulnerable from the risks that we face as individuals and families – remains as compelling now as it was in in the first half of the twentieth century when Western countries embarked on widespread social reform.

During World War II, even when Australia was under the threat of invasion, the Curtin government passed key legislation to extend and strengthen the welfare state – building on institutions that had been created after the depressions of the 1890s and the 1930s. These programs reshaped Australian society. They were courageous and visionary, showed foresight and involved ambitious reforms, even when undertaken during a period of such profound and menacing uncertainty.

Similar vision and foresight are required today, in another period of extreme uncertainty, especially given that we are facing a $1 trillion of debt, an ageing society and a sustained slowdown in productivity.

With an ageing demographic, enormous pressure will be put on the two largest government programs – health care and the age pension. The growth rate in the costs of health care, aged care and government services exceeds inflation and will likely continue to do so. In addition, the welfare state is expanding in areas such as disability and aged care, and providing support to people affected by structural economic changes from globalisation, technological change and rise of insecure work.

I want to make a tangential point that it’s worth observing what the role of government has been over the very long term. The size of government over most of human history, going back to the Sumerians, Romans and Ancient Chinese dynasties was around 5-10 per cent of GDP. Not always, but that was the typical size of government. The first infrastructure that governments tended to involve themselves in was water infrastructure to make agriculture more efficient, then roads. They invested in administration, they had to spend money to keep the palace well fed. Then there was the army. Right across history, in a lot of the graphs that you see, it was only at times of major war that government, as a share of GDP, rose above 10 per cent for extended periods of time.

What we see in the post-World War II period is that government expanded dramatically. The UK has by far the best data. If you look at the UK over the last 350 years, what I call “Roman Empire government” was pretty much all there was, up until around World War II. The welfare state, in terms of not just mandating participation and insurance, but subsidising it, was then around 1 or 2 per cent of GDP. It was after World War II that, in the UK, US and Australia, the size of government expanded dramatically, and it was largely social insurance that drove that. It was a positive development in my view, but we also need to acknowledge that it’s unlikely that we’re going to be able to go through another period where we can expand the size of government by a further 10, 15 or 20 per cent of GDP.

Government might increase a bit as we age, and as we experience cost pressures, and as people expect the government to do more. But I believe it’s going to be a different kind of step change than the one we experienced in the period 1945 to 1980. The other interesting point to note, is that at a very high level in most OECD countries, albeit at different levels, between 1980 and 2020, the size of government as a share of GDP, across all levels, hasn’t moved a lot across most OECD countries. That is important because the welfare state is now so big that we have to understand that broader context to understand how we reform it.

The question is: “How can we do more without a significant increase in taxes as a share of GDP?”

I believe that we can improve their performance not by cost cutting but by achieving better long-term outcomes for citizens. We can deliver better services at a lower cost by implementing a range of practical, innovative strategies.

Risk management needs to play a bigger role. Prevention rather than cure and a greater focus on long-term, whole-of-life investing rather than a focus on the immediate budget constraints. Treating the welfare state as social insurance can encourage risk reduction, better risk management and a focus on well-defined outcomes.

Policy ideas

The first benefit of an insurance mindset will be a greater focus on well-defined, long-term outcomes. Outcomes sound obvious but, as a public servant, I remember in the Department of Finance, we had the inputs, outputs, outcomes frameworks and many other frameworks since. Outcomes are often difficult to specify, let alone measure. It’s extremely difficult, in the life of a person, if you’re trying to have a long-term impact on that person to assess the impact of a particular social policy intervention. There are so many other things happening to them. We know that randomised trials are one way, but they’re difficult to implement in many contexts.

Long-term outcomes

The commercial insurance contract links the payment of a premium with a promise of compensation to be paid in defined circumstances. A well-designed social insurance scheme should activate investment from a massive communal pool to individuals that reflects each person’s needs and preferences. A social insurance scheme should also have a long-term frame. Payments based on a redistributive frame are typically based on a point-in-time assessment of need.

Institutions focused on universal service delivery can sometimes focus on outputs delivered within the confines of short-term budget constraints. In contrast, well-designed social insurance investments can better reflect the long-term needs and opportunities of vulnerable people.

In some areas, universal service is the correct focus, such as vaccination, access to GPs and childcare. In other areas of service delivery, particularly where people have complex and long-lasting requirements, bespoke packages will be more suitable. Those packages should reflect each person’s needs and, as much as possible, their individual priorities. The NDIS is a good example of this – and it will be even more so after Labor has fixed it!

Markets

These can be a powerful tool for eliciting preferences and priorities. In recent times, Australia has experimented with a range of market mechanisms in the delivery of social services. Some of these have suffered from being too complex (offering recipients too many choices through too many providers) and overly complicated and with ineffective governance. We also find many situations where there are so many providers that it’s impossible for the kinds of people who are receiving benefits to make any sense of what choices there are. Many people opt for defaults, no matter what information is put in front of them.

Many of the markets that we have set up also have extremely complicated governance, that often turns out to be ineffective. So, we try and set up markets where we evaluate all of these various providers, but then we end up putting all of our effort into trying to deal with issues like cherry picking and fraud.

But well-designed markets can work. In a pilot that I worked on as a State MP, we established a market for bus transport for children attending a school or autism in Melbourne. Some children were travelling over 90 minutes each way. Many had to change buses mid-route, which was stressful for many students. We worked with world-leading market design experts at Melbourne University and CalTech and, ultimately, established a market that halved travel time, capped travel at one hour each way and removed all bus changes. And for the same cost.

To me, this is a small example of how we can use markets to better reflect people’s preferences, because some parents wish their child to be picked up at home, some didn’t need that. With an extended version of what I talked about; you could have flexibility as to when students were picked up; some parents could have their child stay at school if they were working. It’s a model that could be made more and more sophisticated, could be rolled out more widely, and could be rolled out in schools right across regional and outer suburban areas. It’s a model that could be used to provide transport to train stations for commuters.

Navigators

Empowering recipients is key to effectively enabling greater choice. This can be achieved through a greater investment in navigators and case managers, will enable them to direct scarce resources to what they desire most.

Risk mitigation

A greater investment in risk reduction and mitigation would also have excellent returns, for individuals and in relation to society wide disasters. The Congressional budget office in the United States did a broad examination of schemes that spent money up front in a precautionary manner, and found that many of them were value for money, but not all of them. So, I would certainly say that you need business cases in those situations. As the nineteenth century poem stated: “Better put a strong fence ‘round the top of the cliff than an ambulance down in the valley.”

Shifting risk

Shifting risk can be a game-changer that doesn’t bust the budget. HECS is an Australian innovation that proves the point. It has opened up higher education opportunities for many. All of the studies that have compared HECS type schemes to schemes such as the one in the US, where people are given access to loans or subsidised loans, HECS provides much greater access and at a fraction of the administrative cost. Particularly for people who are disadvantaged. Income contingent loans could be applied in many other areas including social housing and medical expenses.

The same approach could also be used when government partners with industry, particularly in supporting cutting edge innovation – taking an equity stake and de-risking projects rather than simply offering grants.

As a state MP, I worked on a scheme to fund women’s change rooms. The existing grants program disbursed $14m and was a very worthwhile scheme. But it helped only a fraction of the clubs needing new facilities given the burgeoning demand for new women’s and girls teams across a range of sports. The new scheme offered an additional option. Instead of grants, under the new additional scheme, government partnered with sports clubs and local government – backing low interest loans at the government’s AAA rating and assuming some of the project risk. This saw a 10-fold increase in how many projects the government could fund for a given public investment.

Sustainability

Finally, an insurance and risk management approach offers the best way to rigorously and transparently measure and report on the long-term sustainability of our key social insurance institutions. Sweden’s age pension system is a case in point; they adjust at the margin so that there’s intergenerational fairness. When demographic projections come in and turn out to be different from what was expected, each generation, both the current taxpayers paying a Pay As You Go pension, but also the people receiving it, share some of the gain or cost.

In conclusion, our welfare state is one of society’s greatest and most civilising achievements. It arose from insurance schemes and was motivated, in part, by risk management principles. By returning to those origins, we will be able to improve the outcomes for the most vulnerable in our society and help preserve the public institutions that have served us so well.