Australian Priority Investment Approach to Welfare

Speech Location:

National Press Club, Canberra

*** CHECK AGAINST DELIVERY ***

I thought I might start this speech, and perhaps even finish it, in 1987. That was the first time I can recall paying attention to a Press Club speech; that was a speech given by Paul Keating as Treasurer in 1987, in his very personal style.

That was my first year at University - and like many other people, that was the first time I really started learning and thinking, in a formal sense, about the role of government, the nature and size of the welfare state.

I look back on 1987, as many people here would know - music was very bad, haircuts were worse – mine was particularly bad at the time.

After, and in the wake of Reaganomics; every Politics 101 course was completely obsessed with this theoretical question of what should be the proper size of government, like there was some magical answer to that question.

I also distinctly recall that every single exam question invariably ended in the word ‘discuss’.

Should the welfare state be bigger or smaller – DISCUSS.

To achieve more equality should there be more or less income redistribution – DISCUSS.

For a long time, I must confess to being one of those who found the bigger or smaller government debate utterly fascinating.

In fact, it wasn’t until I actually had to come to deal with public finances for real that I found that the theory, or ideas around bigger/smaller government to be utterly unconvincing, and not of much use or merit whatsoever in real world government.

Mind you, I would confess to you all that at university I was quite taken with the ethics of liberty notion, that all tax was theft - but I found that theory somewhat less attractive almost as soon as I became a State Treasurer.

So we all grow up a little I guess.

The notion that some or other ideology can prove that there’s some perfect or right answer to the question of how big the welfare state should be - seems to me, at least now many years later, to mistake a question of context for a question of theory.

It also seems to incorporate some strange assumptions about the purpose of welfare.

The issue of welfare’s purpose, or true purpose, is one that I’ll return to shortly.

But on the issue of context - in the real world, the size of the welfare state question requires, first and foremost, a consideration of the prevailing circumstances.

What should be spent on welfare is surely a question answered fundamentally by recourse to real world conditions; rather than answered by opening a text book.

The context we faced as a government in 2013 was inheriting a line of structural deficits.

That was a mess, but it was not a mess of our own making – simply put, Labor had imbedded future structural spending which was substantially in excess of realistically deliverable revenue.

Today’s welfare spend is around $160 billion a year.

And as a matter of scale, or comparison, that represents 80 per cent of all individual income tax raised in Australia. So $196 billion in income tax, 80 per cent of that is, in a practical sense, allocated straight to welfare.

Without further restraint, the welfare bill by 2026 will expand out to $277 billion.

Even more astonishing, and going to that point of embedding the structural spending beyond which revenue we could realistically get - when Labor built the stimulus welfare payments into the Budget in 2008-09, the yearly welfare bill at that point climbed to over 100 per cent of the personal income tax take. A very large amount indeed.

Under Labor, the growth rate of social security and welfare spending was almost double the growth rate of revenue.

Social Services is, as was noted, simultaneously the largest area of government spending and the fastest growing, with growth outstripping GDP and outstripping CPI.

Given the context depicted here – a theory or ideology that says to anyone in government that we should have a faster growing, or even bigger welfare system just makes no practical sense.

Continuing to borrow money to fund today’s welfare system growth simply loads on debt to be repaid by our children, and that option would see young Australians, in effect, paying for the welfare system of their own time in the future and retrospectively paying for the welfare system of our time.

Some people will avoid attention being drawn to this intergenerational unfairness but I have not heard a single person seriously argue that we should simply sustain the trajectory of welfare growth that we have by perpetually borrowing more and more money.

That approach, in my observation, would be a fundamental breach of the democratic compact, to leave the nation to the next generation in better shape than we inherited.

Another approach advocated by some is that we should simply increase taxes on working and investing Australians outside the welfare system to pay for the unrestrained growth of the welfare system.

The major problem I think, with the more tax approach, arises from a combination of several real world features of the system:

  1. We have a broad and a very generous welfare system.
  2. We have a very progressive income tax system.
  3. We have a high income tax-free threshold, by international comparisons.
  4. And critically, we have very high and generous income free areas. These are the amounts various welfare recipients can earn before they lose a dollar of welfare.

Those four things combine, in a unique way in Australia, to create the situation where 48 per cent of households pay no net tax.

The situation where you have just on half, and potentially more households drawing from the system than contributing to it is, again in my observation, like a snake eating its own tail. That is to say that it does not work so well after about half way.

If real world circumstances demonstrate the need to take reasonable efforts to restrain welfare expenditure growth, than those same circumstances must also mean that there is an absolute imperative to ensure we spend welfare dollars effectively.

We must get the very best outcomes for the very significant amounts of welfare money that we do spend.

As a confessional, I think it would be fair to say that governments of all stripes have generally not been very good at measuring the effectiveness of the welfare spend.

The result is that very significant parts of the welfare system have become totally ineffective at producing better outcomes for the people who become trapped in the system.

Part of the problem has been that no previous government, not until this one, has made the concerted effort necessary to amass, and collate, and construct the data necessary to measure outcomes.

Another problem is the persistence of the 1980s ideological fixations.

Applying ideological measures means you often measure the wrong thing, and what has happened is that, Government’s very often mistake effort for outcomes.

The ideological proposition that there is some wrong or right level of inequality is a good case in point.

Whether you hold a view that inequality is massive in Australia, or modest, or getting better or worse, or by international comparisons, good or bad, the real issue is that inequality is a measure of difference, not a measure of comparative wellbeing.

A decreased level of inequality, measured by some change to a complicated income co-efficient, tells us precisely nothing about whether any change has improved a single individual life. Not a thing.

Government can transfer money from one group to another, but for too long the measurement of success has stopped at the mere fact of the redistribution.

In the real world, the mere transfer of money from one group to another is not a good enough reason to hang out the ‘mission accomplished’ banner. And that’s what has happened for decades.

More spending itself is no guarantee that key groups have had their lives improved in any material, ongoing or significant way.

Say for instance that it can be shown that cash welfare payments, expended too frequently on alcohol in certain places, at certain times, while making inequality coefficient measures better on paper, actually make human lives worse.

That is the real world problem that is a problem that the Human Services Minister, Alan Tudge is trying to fix, with strong early signs of success with the cashless debit card, the healthy welfare card.

I’m not suggesting that all welfare spending has counter intuitive results, of course not.

But data now available, and that we’re releasing today, demonstrates that there are too many instances where spending is failing to produce substantial improvement to Australian lives.

There are simply too many instances to be ignored:

  • Where young people enter the welfare system and remain inside it for their entire lives.
  • Where recipients can receive more in welfare than the minimum wage without being subject to ongoing obligations to become work ready.
  • Where payments require no mutual obligation to find and hold even a modest number of working hours.

The sustainability of the welfare system is critically important to the next generation of Australians, but true welfare reform is much more than making budget numbers sustainable.

Equally, the true purpose of welfare is not to achieve these ideological targets around inequality or the size of government.

The real task is to continuously engage in an honest, evidence-based assessment as to when an existing welfare spending approach improves an individual’s prospects for a better life.

A life made more meaningful by employment, by community contribution and through self-reliance.

And measure when it does not.

There is nothing particularly progressive about parts of a system which fail to place fair and firm obligations on capable individuals to assume greater responsibility for their own lives.

There is nothing morally superior about welfare structures that are passively allocating money in a way that corrodes the recipient’s chances of experiencing the meaning, the engagement and the purpose that work brings into our lives.

This brings me to the Priority Investment Approach, as a very significant, maybe close to revolutionary, new direction in welfare reform.

I’ll go back quickly, on 9 May 1974, the producer and critic Jon Landau saw Bruce Springsteen at the Harvard Square and said he had seen the future of Rock and Roll.

I grant you that ‘Baseline Valuation Report’ is not a catchy title like ‘Born to Run’, but I do earnestly believe we have real future insight here as to how data can drive policy development.

In 2015/16, this was part of the plan by my predecessor Scott Morrison, who I’d like to note today, part of a plan that fits in with other things that we are doing which was started well before my time; we committed $33 million to pursue this Priority Investment Approach to Welfare.

We have constructed a purpose-built data system. It collates all the welfare information that we have collected across Australia, over the last 15 years.

It allows us to then apply the types of algorithmic and actuarial analysis used in insurance industries.

Through this modelling we can predict the likely movements for target groups on, off and in between welfare payments, and calculate welfare expenditure for groups over the lifetime of the system.

Applying all of this analysis to all 24 million Australians, and I must make the clear point, that’s both those presently in the welfare system, but also the ones the analysis predicts will come into the welfare system, the PwC report, and thank you to those members from PwC who are here for all your hard work, the report shows that we face a total estimated future lifetime welfare cost of the present Australian population of $4.8 trillion.

That is a very large figure, and no doubt it will receive considerable attention.

But the really important purpose of the new data system is that we can be provided with great detail about quite small groups, and so we can use the data to test policy approaches to see whether they actually contribute to full, purposeful and self-reliant lives.

We have set three goals for a new direction in welfare policy, which are to:

  • Identify those at high risk of long term welfare dependency and help them find employment;
  • Secondly, identify and reduce the risks of welfare reliance crossing generations; and
  • Thirdly, ensure the long term financial sustainability of the welfare system.

The first use of the data has allowed us to identify and focus attention on three specific groups who have particularly poor long-term outcomes in the system.

Young carers, young parents and, surprisingly, a group of young students.

Looking first at young carers:

And I meet now, through this job, very impressive young Australians all the time, and as I meet them, and I know what this data system tells me are the risks for their future, I have to say to you I think there is a moral, not economic imperative that we help this quite small group of young carers.

The reliance on welfare payments by a carer, structurally speaking, is meant to only last as long as the care relationship. The issue and the challenges are with the recipient of the care, not with the person providing the care.

The data we have now assembled shows that in any year over the next 70 years a minimum of 40 per cent of these 11,000 young Australians under 25 will be expected to access income support payments.

Even more concerningly, a quite substantial number of these young people, 16 per cent, that’s 1,800 people, will be accessing income support in each year for the rest of their lives.

On average, the 11,000 young carers are expected to be on income support in 43 separate years over their future lifetime.

The baseline valuation estimates the total future lifetime cost to the welfare system, of this small group of 11,000 people, is $5.2 billion.

And that $5.2 billion says nothing about the human cost.

That a young carer’s commitment to provide care for others leads to their own long-term welfare dependence, means we effectively, all of us but particularly government, maintain a system which helps parents at the cost of lost opportunities and unrealised potential for their children.

We also know that the number of young carers in the last decade has trebled.

Given that young carers sacrifice to help provide for those in need, and we have the knowledge that their own future is likely to be one of welfare dependence. This is a moral as I say, not economic imperative.

So what can we do?

The Coalition has already started, based on instinct, around what we think is the right approach.

We’ve started the Young Carer Bursary Programme, a $3.5 million investment designed to improve the educational outcomes of young carers by helping them complete their schooling.

It seems to make sense, that’s intuitive, but the important point is, we can now test this policy. We can exact this actuarial discipline on ourselves as a government, and see if it works in a demonstrable and clear way.

The Priority Investment Approach will be used to evaluate initiatives like the Young Carer Bursary and other initiatives across government, and so grow successful programmes and shrink the unsuccessful ones with the aim of turning trickles of people out of the system into streams.

The second group that emerges as particularly vulnerable are young Australians receiving the Parenting Payment.

So here we’re looking at 4,370 young parents under 18. They receive Parenting Payment, about 77 per cent of these 4,370 are single.

Our new data system predicts that this small group are very likely to have further children, remain on Parenting Payment for very long periods of time.

Looking back, which is how the data structures its predictive algorithms, from 2002, 1,580 young Australians continuously received parenting payments for 13 years. All had at least two children, with two-thirds having four or more children.

The baseline valuation predicts that for the similar group going forward that the total future lifetime welfare cost for the 4,370 young parents under 18 today is $2.4 billion.

If nothing changes, 12 per cent of these young parents will access income support in each year for the rest of their lives.

On average, these 4,370 young parents are expected to access income support in 45 separate years over their future lifetime.

In addition, there is a truly terrible intergenerational welfare dynamic happening here.

Eighty per cent of the young mothers entering the welfare system for the first time had a parent or guardian who also received income support during their upbringing.

Reducing welfare dependence among this group could make a massive difference to their own lives, but also we know statistically, it can change and improve the prospects for children who are otherwise are simply full of potential.

Findings like these sharpen our picture of welfare reliance crossing generations, and we need to use the new data tool to understand why the present structure of the system has been so ineffective for this and for other groups.

The third group is young students.

The Report shows there were nearly 400,000 Australians on student payments in the last financial year. They received a total of $3.3 billion in that year.

Instinctively, you, like myself would have thought as most Australians would do, that this group who receives $3.3 billion to study should represent some of the better results in terms of transition to future employment.

But the data shows some really quite disappointing results.

Overall, in any year over the next 60 years close to 30 per cent of those 400,000 students who receive payments today will be in the welfare system.

There is then a smaller group, we’ve identified about 6,600 students who received a welfare payment to study, and they did so between 2003 and 2011, they then transitioned onto a working age payment, so an unemployment benefit of some type, and they stayed there for more than 12 months.

If nothing changes, it is estimated that for this group of 6,600 students, more than 9 per cent of these will access income support in each year for the rest of their lives.

On average, this group is expected to access income support in 37 separate years over their lifetime.

Looking at these three examples, what I would put to you, and what our Government has determined is this, is that more money allocated to existing frameworks is not the answer.

For too many people in the system money flows, nothing changes, and lives are not improving.

New approaches, new structures, they have to be tried to improve the lives of groups who become trapped in the welfare system.

Our first critical response, and part of today’s announcement as well as the release of the Baseline Valuation Report, is to create a $96 million Try, Test and Learn Fund.

The Coalition Government, the Turnbull Government will provide the funding, we will provide the data, we will provide the tailored metrics against which success and failure will be measured.

The Fund will be BYO ideas.

There are any people in this room, from groups in the community who will provide those ideas.

Stakeholders, academics, state and territory governments, anyone in the non-government sector who thinks they can produce a better result, will be able to put forward their ideas.

Chosen plans will be subject to a simple basic metric of performance; does the plan improve lives by increasing self-reliance over time?

The Fund acknowledges that government simply does not have the answers to all problems. But we can start to provide, through the expenditure of targeted funds, the tools and the discipline to test new plans to see if they actually can provide real answers.

Rigorous evaluation will be the hallmark of the system. We’ll ensure that we do not continue down the often trod path of spending money in an area merely because it instinctively or emotively appears to be useful.

That is a path that has led to the situation we are currently at in the welfare system, where funding for programs that may not be having an effect is continued for too long.

After an initial period of operation of the Fund, we will be in a position to set targets for reducing the lifetime costs of the identified groups already inside the system and set targets to reduce the number of people in any of the priority groups reliant on welfare in the future.

This is a very fundamental change to social policy.

Initially, we will be looking to target those groups identified, but it won’t stop there, future rounds can and will focus on other groups that we’re looking at:

  • Recent exits from the welfare system.
  • Older people entering carer payment and exiting carer payments.
  • Parent’s whose youngest child is approaching the age that parenting payment ends.

But as well as this policy targeting approach, the data we now have provides very strong reasons to question the effectiveness of some of the fundamental overarching architecture of our welfare system.

As well as targeting specific groups, the data that we have amassed can help devise better, simpler overarching structures that could help achieve better outcomes and improve lives across groups.

Last year, here at the National Press Club, my predecessor Scott Morrison released the McClure Review into Australia’s Welfare System.

That Report concluded our welfare architecture to be, simply put, excruciatingly and unnecessarily complicated.

The effect is that it is hard for people inside the system to understand it and very hard for people outside and in charge of the system to measure what is inside it and what is going on inside it.

For anyone who doubts how crazily complicated the present system is, have a look at Patrick’s most simplistic diagrammatic representation of the Australian welfare system.

This is just the mud map, it’s as simple as it gets.

That maddening complication frustrates the development of simple, understandable and swiftly enforceable incentives and responses designed to encourage behaviours that transition people into employment and discourage behaviours that lead to dependency.

Presently, people in very similar circumstances can be subject to very different payment regimes, and so have very different requirements and different levels of mutual obligation.

Astonishingly, the present structure of the working age payment system sees that 400,000 people with working age payments have no mutual obligation whatsoever to even try and search for appropriate work.

Mutual obligation must be cornerstone of any working age payment redesign and the central obligation should obviously be the requirement to prepare for, to search for and to accept work.

But mutual obligation can be very successfully applied further to greater effect, as we have seen, with the obligation to vaccinate children.

If it works there, why could mutual obligation not also extend in appropriate circumstances to:

  • An obligation to refrain from excessive alcohol or from illicit drug use, where evidence clearly shows it creates barriers to employment.
  • To obligations to turn up in a timely manner to key work appointments.
  • To pay debts owed to the taxpayer.
  • To ensure children attend school.

If there was ever a time to consolidate the complicated mess of 16 different types of working age payments into a simpler, fairer system with more thorough and consistent mutual obligations then that time would be when we have the data available to help us design better structures, rules and systems. And the data is now available.

So by way of conclusion, I started this address in 1987 and looking back, it occurs to me that we all have formative years and whether they go smoothly or not is often due to matters totally outside of our control.

Coming from a family of modest means, I nevertheless had everything you could want from an Australian family.

In 1987, I was free to pursue studies without encumbrance of any type.

Since then, like all Australians, and many in this room of my generation, I have seen my parents get old and get all the serious health problems that come with that age.

So it’s natural in a position like mine when I see and meet people to think back about how different things might have been. What if I had to face the challenges that I face with my parents’ health when I was a teenager, instead of when I was in my 40s?

The hope for a young Australian facing challenging family circumstances today will be that the system is going to provide for them with immediate support, but also surely they should expect that the same system will not just set them and forget them.

Surely they expect that as well as welfare money given in the here and now, that support is going to be thoughtfully structured in a way designed to help them cope, not just in the present, but also in a way that maximises their future opportunities for self-reliance.

With the information we have now, with all the information that’s assembled and a will from this Government to use it, we have a real chance to provide welfare assistance to individual Australians in a way that measures outcomes not by ideology or emotion or by amounts of money alone.

But by assessing the long term effects on the real lives of the people we insist we are trying to help.

Thank you very much.