From Basic Income to Poor Law and Back Again
Part 4: Designing a Viable Basic Income
This is the final in my four part series on basic income. In Part 3, I highlighted potential impacts of different basic income options on the labour market and social security system. This demonstrated that basic income schemes can have adverse unintended consequences, especially for low paid workers. In this final Part 4, I examine three possible responses to these dilemmas:
(1) a means-tested Basic Living Income;
(2) a less strictly means-tested ‘life cycle’ Living Income; and
(3) a set of smaller Supplements to meet particular costs.
These could be combined to form a two-tier Basic Income scheme with a base rate of payment to meet general living costs and a supplementary tier to meet additional costs.
(1) A Basic Living Income
The most comprehensive proposal for a Basic Living Income in Australia comes from the Australian Council of Social Service (ACOSS). It advocates replacing the pension and allowance payments for people of working age with a single basic payment (with different rates for singles and couples).
The payment level would be set to meet the full minimum living costs of a single adult or couple. In setting payment rates, the government would be advised by a statutory expert Social Security Commission which would undertake research on living costs for people with low incomes and report regularly to Parliament. The Age Pension and Family payments would remain in place. Rates of payment would be indexed to wage movements and consumer prices, and the payment would be income-tested.
In this way, the present income support system for adults of working age would be converted from a hodge-podge of payments with different rates and eligibility requirements into a form of Basic Living Income or minimum income guarantee.
This is not a new idea. When the Whitlam government was elected, Social Security Minister Hayden announced that:
‘The New Labour Government hopes to completely scrap the present confusing system of pension and social security benefits and replace it with a more simply administered and easily understood system of guaranteed income.’ (Hayden 1973)
Reciprocity and conditionality
A key issue is the extent to which a Basic Living Income should be conditional. Should activity requirements such as job search or training apply? On what basis people should be exempted from them (e.g. disability or caring roles)?
This raises issues of reciprocity: what (if anything) should people be expected to do in return for a living income from the State?
Prominent British welfare theorist, Richard Titmuss explained in 1960:
“Modern social welfare has really to be thought of as ‘help given to the stranger, not to the person who by reason of personal bond commands it without asking’ [citing Wilensky & Lebeaux (1958)]. It has therefore, to be formally organised, to be administered by strangers, and to be paid for collectively by strangers.” (In Abel-Smith & Titmuss 1987, p85)
An impartial, rights-based approach is what distinguishes social security from charity. The entitlements to this ‘support from strangers’ rest on a combination of political and institutional power (the balance of interests represented in Parliament) and a supportive ideology based on social solidarity and reciprocity among recipients and taxpayers – who are ultimately the same people. We see this every year in the ebb and flow of federal budget debate on social security.
Conditionality has many variants, from the Poor Law system of punishment for the ‘undeserving’ (virtually everyone except older people and those with a disability) to the Nordic ‘active line’ in which employed and unemployed workers are expected to ‘do your duty and demand your rights’ (Dølvik et al 2015). Expectations of what government will provide range from unpaid work for benefits in Australia’s ‘Work for the Dole’ scheme to a right to employment and training programs, as well as benefits, in Denmark (Kvist 2007).
The purpose of activity requirements has been to promote self-reliance through job search and participation in employment and training programs. The risk, as discussed in Part 2 is that those requirements become so demanding and so out of touch with people’s actual employment prospects that they become a form of punishment.
In recent years, a ‘new paternalism’ has taken hold in social security policy, which takes activity requirements a step further. It aims to curb what it calls ‘welfare dependency’ by regulating the social behaviour of recipients of income support. New paternalists call for drug tests and compulsory treatment, good parenting requirements and restrictions on how people spend benefits such as ‘income management’ of benefits by Centrelink or ‘cashless welfare cards’ that ban spending on drugs or alcohol.
This is a slippery slope towards a new category of second-class citizen. People receiving working–age social security payments are increasingly expected not only to make reasonable efforts to attain financial self-reliance, but also to accept invasions of privacy and behavioral restrictions that in the past may only applied to people convicted by the courts of serious offences. Predictably and regrettably, in Australia, the first group singled out for this treatment have been people in Aboriginal and Torres Strait Island communities, who weren’t even granted citizenship until about 50 years ago.
A Basic Living Income could be paid unconditionally or the range of justifications for payment could be broadened as Atkinson (1996) proposed with his ‘Participation Income’, to include unemployment, disability, education and training, and parenting and other caring roles. The Australian social security system does this already in its own parsimonious way. Parenting Payment supports the primary carers of children, but only for preschoolers. Youth Allowance and Austudy Payment support full-time study, but are deliberately paid at a lower rate than Newstart Allowance.
To adapt the system to present-day conditions and needs, we could broaden these definitions, extend eligibility to more people and pay all at the same rate based on need.
(2) Life cycle payments
In Part 3, I discussed how supplementing low pay and insecure work with a small universal payment could have adverse unintended effects. Yet unemployment, low pay and inadequate or insecure paid working hours are not the only income risks people face through life. In addition to a Basic Living Income, there is a role for ‘life cycle’ payments to cushion people’s incomes as they undergo other changes in their lives that put their incomes at risk.
The two biggest life-course income risks most people face are the costs of children and supporting ourselves in old age. Australia’s social security system recognizes these risks through family payments (Family Tax Benefits A and B) and the Age Pension. As it is recognized that most of us face these risks and these payments are not income-tested as strictly as income support for people of working age. In 2014, 60% of families with a child under 16 years received a family payment and in 2016, 66% of people over 64 years received an Age Pension, compared with just 18% of people of working age receiving income support in 2015.
This chink in the armour of Australia’s defenses against so-called ‘middle class welfare’ is well justified (Henderson & Spies-Butcher 2017). Most countries provide some form of social insurance against these risks. Whether these payments should be more loosely income-tested depends on our priorities for the federal budget. If we ‘target’ them to such a degree that less than half the eligible population is entitled to full or part-payment (and we may be approaching this benchmark with FTB, with the tightening of income tests after 2014), then we have probably gone too far.
There is also a strong case for smaller supplementary payments to meet non-discretionary costs above the basic living expenses that we all face. Examples include Rent Assistance (for low income households renting privately), Carer Allowance (to assist carers of people with disabilities with the costs of care), and proposals for a costs of disability allowance to assist with the direct costs of a disability (ACOSS 2014).
Since these costs exist whether a person is unemployed or in paid work, there is a strong (and publicly supportable) argument for broadly-based public support to meet them. Another advantage of supplements is that they are less likely to be viewed by employers, governments or Fair Work Commission as de facto wage subsidies because not all employees would receive them. This is one way around the risk of triggering a ‘race to the bottom’ on wages.
Designing a basic living income
A guarantee of a Basic Living Income to provide income security for all is a hallmark of a good society. Australia’s current social security system fails to achieve this.
On the face of it, the Australian system is a long way removed from proposals for universal Basic Income. Indeed, it is more strictly income-tested than any other wealthy OECD country (Whiteford 2016). Yet the design of Australia’s system, being based on need, means that it comes closer to Basic Income principles than other wealthy countries where payments are based on the insurance principle and linked to the previous income of the recipient. This is important for those at greatest risk of poverty, including people with disabilities and women who have spent years out of paid work while caring for children.
Support for a Basic Income is weakest in those European countries with robust social insurance systems such as the Nordic countries, as shown in Figure 2. In countries with ‘Beveridgean’ social security systems like the United Kingdom, people are more supportive of the idea of a basic minimum income for all. This suggests that popular support for a need-based Basic Living Income might be stronger in Australia than in most OECD countries, but that our preference for means-testing means that universal payments are more likely to be resisted.
Figure 2: Support in Europe for a Universal Basic Income to replace existing social security payments
Source: Fitzgerald R (2017)
At its best, the Australian social policy tradition offers support based on need and the redistribution of resources to reduce poverty and social inequality. At its worst, it applies an outdated Poor Law distinction between ‘deserving’ and ‘undeserving’ recipients and punishes the undeserving with increasingly intrusive and unrealistic activity requirements. Our ‘liberal welfare regime’ (Esping Andersen 1999) means that adequate income support is too often regarded as a last resort to be offered reluctantly and sparingly, to discourage so-called ‘welfare dependency’.
John Quiggin’s argument that we should give more weight to the ‘Basic’ than the ‘Universal’ in Basic Income schemes makes sense. A Basic Living Income would give priority to those in most need of support, and reduce the risk that they end up with lower social security payments or lower wages as an unintended consequence of reform. This suggests two directions for reform:
First, we should do away with the distinction between pension and allowance payments for people of working age and replace both with a means-tested Basic Living Income scheme.
Secondly, the system should reach beyond those at immediate risk of poverty to assist low and middle income earners to negotiate life-cycle risks such as the costs of children and retirement, and provide supplements to meet unavoidable costs above and beyond the basic expenses faced by everyone. How strictly these lifecycle and supplemental payments should be income-tested is a pragmatic question of budget priorities.
The social minimum is about much more than social security legislation. Income security for all requires full employment, decent hourly wages, adequate and secure paid working hours, and a strong foundation of support services including employment assistance, lifelong education and training and child and elder care. A decent basic income, full employment and a welfare state are complementary, each supporting the other.
This four part series is written based on the presentation, ‘From basic income to poor law and back again: can a UBI break the Gordian Knot between social security and waged labour?’, by Peter Davidson at the Australian Social Policy Conference at UNSW on 27 September 2017.
It is serialised on this site, and on the Austaxpolicy blog, at: Davidson, Peter (2018), From Basic Income to Poor Law and Back Again – Part 4: Designing a Viable Basic Income, Austaxpolicy: Tax and Transfer Policy Blog, 22 March 2018, Available from: http://www.austaxpolicy.com/basic-income-poor-law-back-part-4-designing-viable-basic-income/
From Basic Income to Poor Law and Back Again
Part 3: Renewing the Social Minimum
In part 1 of this series on Basic Income, we travelled back in time to England on the eve of the industrial revolution in search of the first Basic Income scheme. When capitalism supplanted the traditional master-and-servant system of pre-industrial England, the Speenhamland system of local benefits for rural workers was replaced by the brutalities of the New Poor Law that denied benefits to ‘able-bodied’ unemployed people so that the factories had a ready supply of labour.
In the 20th century, the labour movement and social reformers learned from the past that under capitalism, the labour market, politics and social welfare are intertwined. They built a ‘welfare state’ on a foundation of universal suffrage, full employment, labour regulation, universal social services, and social security for people lacking enough income to live decently.
In part 2, we identified challenges to the modern welfare state including precarious employment, attempts to wind back social spending, and the increasingly harsh treatment of people relying on working-age social security payments which recalls the Poor Law distinction between ‘deserving and undeserving’ poor. This has led to new questions about whether, and how, a Basic Income scheme might be part of the solution.
This piece (Part 3) examines the structure of the ‘social minimum’ (minimum incomes) in Australia, and explores the implications of replacing parts of this system with different kinds of Basic Income.
The three pillars of the social minimum
Income support is only one element of the ‘social minimum’ – the set of social guarantees that underpin income security in wealthy capitalist societies. The three pillars of a decent minimum income are the family, labour market regulation, and the welfare system, broadly defined.
Figure 1 shows trends in key components of the social minimum income in Australia.
Figure 1: Wages, benefits and pensions for a single adult ($ per week, adjusted for inflation)
Sources: ACOSS (2017), OECD StatExtracts; Department of Social Services
Note: Minimum wage for a fulltime worker, excluding overtime. Average weekly ordinary-time earnings in main job for men and women. Family payments for two children (one of primary school-age, the other of preschool age), including maximum Rent Assistance.
- Newstart Allowance (unemployment benefit) is very low (currently $38 a day) and has not increased above inflation for over 20 years. The last ‘real’ increase was a $2 a week rise from the Keating government in 1994.
2.Minimum wages have barely increased in real terms for two decades, and there is a reasonably consistent relationship between them and Newstart Allowance rates.
- Consistent with Poor Law principles, allowances for those deemed ‘able to work’ are much lower than pensions for those deemed ‘unable to work’.
- Apart from pensions (which are indexed to wages), the social minimum is falling behind wider community living standards, proxied here by the average full-time wage. That is, the benefits of rising productivity and living standards have largely been denied to unemployed people and minimum wage-earners for two decades. Their living standards were effectively frozen in an era when the internet was in its infancy.
- Family payments for those with low incomes rose in real terms during the 2000s but have since declined. This is due to the removal of their indexation to wages in 2009 and subsequent decisions to freeze maximum rates of payment.
Back to a Basic Income?
A Basic Income scheme alone cannot guarantee that everyone in or out of paid work has a decent income. Critics of universal basic income are right to warn that replacing employment, wage and welfare protections and services with a minimum income guarantee is risky. It risks a repeat of the Speenhamland experience that powerful interests pull the new system down, arguing that it’s too costly and work incentives and the ‘dignity of work’ would be diminished. Another risk is that income protections in the other pillars of the social minimum – especially wages and human services – would be adjusted downwards.
A better question to ask is whether a Basic Income, together with reforms to strengthen the other pillars of the social minimum, could ensure income security for all.
Much depends what kind of Basic Income scheme is introduced. Some options are:
- A ‘living income’ (which people can live on in accordance with general community expectations – that is, above poverty levels) to replace the present social security system. This has two variants – an income-tested payment and a universalone.
- An ‘income supplement’ which is not enough to live on, but supplements wages or social security payments to give people more flexibility to negotiate a more uncertain labour market, help with extra costs (such as a disability or retraining) or to combine employment with other roles such as care. This is usually advocated as a universal payment, or one that at least extends to the majority of income-earners.
At the heart of the contest between different Basic Income models are tensions between adequacy and universalism, and between the obligations and entitlements of citizens.
Either a living income or an income supplement could be conditional (for example, on labour market participation) or unconditional (for example, an entitlement of citizenship). This distinction was drawn by Tony Atkinson, who advocated a ‘Participation Income’.
Tensions between adequacy and universalism arise due to the high cost of a universal scheme providing sufficient income for people to live on. As Martinelli (2017) puts it: ‘an affordable UBI would be inadequate, and an adequate UBI would be unaffordable’.
Whiteford estimates that a universal, unconditional Basic Income set at the pension rate ($21,000 a year for a single adult), with allowances for partners and children, would cost $360 billion a year, compared with the $150 billion cost of existing social security payments. A large increase in tax rates (not only for high income-earners) would be needed if this were funded through the income tax system, even if the Univeral Basic Income replaced the tax-free threshold (Scutella 2004).
On the other hand, a Universal Basic Income costing the same as the current social security and welfare programs would yield a payment of around $6,000 a year, less than half the already inadequate Newstart Allowance ($13,500).
Basic Income options for Australia
Most debate in Australia has focused pragmatically on two options (the shaded segments of Table 1):
- A ‘Basic Living Income’ that replaces existing income support and is means-tested. An example is the ‘common working-age payment’ advocated by a government welfare reviewin 2001.
- A modest ‘Universal Basic Income’that supplements income support and minimum wages and is not means-tested. An example was the ‘Guaranteed Minimum Income’ advocated by the Henderson Poverty Report in 1976.
Table 1: Types of Basic Income
Impact of options on the labour market and welfare system
What is the likely impact of these basic income options on the labour market and welfare system?
Much of the debate over the labour market impact of a Basic Income scheme concentrates on ‘work incentives’ for paid work. As Martinelli (2017) and Bowman et al (2017) point out, this argument has been over-stated. Experimental unconditional Basic Income schemes in the United States and elsewhere have only marginally reduced labour force participation. The myth of the ‘dole bludger’ who would rather do nothing at home than work is just that. The main impacts on paid workforce participation were among people who gave priority to other activities, especially caring and studying.
The more interesting, and more important, labour market impact of a basic income is its effect on wages. This could go either way. If workers have an alternative to working for wages they may drive a harder bargain, in which case wages would rise. Alternately, if workers have access to a wage supplement, employers may drive a harder bargain to capture this subsidy, or the Fair Work Commission may discount minimum wages, in which case wages would fall.
Option 1 (Basic Living Income) is likely to improve pay and conditions for low-skilled work. It might (modestly) reduce workforce participation among unemployed people, especially if unconditional. Low-skilled workers, who are less likely to have substantial family resources or personal savings to fall back on, would have a viable alternative to working in a job they don’t want.
If the Basic Living Income were income tested to moderate its cost, then it is of no immediate benefit to middle and higher income-earners, but would still play a vitalinsurance role for those households. We all face risks such as redundancy, ill health or marital separation.
Option 2 (modest Universal Basic Income) is likely to strengthen pay and conditions for higher-skilled work and may reduce their paid workforce participation slightly (mainly among parents and other carers). This is because even a modest Basic Income would enhance choices for those with significant family support or savings.
However, it would probably reduce wages for low-skilled workers who lack family support or savings. A modest Universal Basic Income would do little to improve their bargaining power. Unless it is built on a solid foundation of robust minimum wages and secure working hours, the subsidy is likely to captured by employers, especially if Fair Work Commission takes account of a major increase in public support when setting minimum wages.
Climbing or sinking?
(Source: Koi Bito Forum)
The end result could be an increase in wage inequality. By supplementing low-paid, insecure work, there is a risk that a Basic Income that is too low to live on could entrench it for people lacking bargaining power.
In Part 4, I will examine three possible responses to the dilemmas identified above: (1) a means-tested Basic Living Income; (2) a less strictly means-tested ‘life cycle’ Living Income and (3) A set of smaller Supplements to meet particular costs. These could be combined to form a two-tier Basic Income scheme with a base rate of payment to meet general living costs and a supplementary tier to meet additional costs.
This four part series is written based on the presentation, ‘From basic income to poor law and back again: can a UBI break the Gordian Knot between social security and waged labour?’, by Peter Davidson at the Australian Social Policy Conference at UNSW on 27 September 2017.
From Basic Income to Poor Law and Back Again
Part 2: Whither the Welfare State?
In the first part of this series, we followed the introduction and abolition of the first ‘Basic Income’ scheme, the Speenhamland system in the United Kingdom in the 19th century. When Britain industrialised, cash benefits were replaced by the ‘New Poor Law’ and the Dickensian workhouse. The conclusion drawn by social reformers was that to end poverty and financial insecurity, they would have to work on a broad front: from industrial regulation to universal suffrage and the construction of a welfare state (social security, education and community services). This strategy was very successful, but now there are concerns that it no longer works and can’t be sustained.
The Australian post-war welfare state was built upon a ‘social minimum’ in two parts.
In the labour market, it included a high minimum hourly wage, full employment and regular working hours (for men), and a high unionisation rate that supported collective bargaining and industrial arbitration. Frank Castles (1996) described this as a ‘wage-earners’ welfare state’.
This was complemented by social protections won through universal suffrage and political campaigning. They included free public education, (mostly) free public health services, a robust public Vocational Education Training (VET) system, the Age Pension, an unemployment benefit safety net backed by a public employment service, and family payments to supplement the minimum wage and reduce child poverty (both for families in & out of paid work).
It’s now argued that these social protections (at least for people of working age) cannot be sustained as the neo-liberal form of capitalism has removed trade barriers, restricted public spending, and brought intense competitive pressure to bear on the labour market so that a regular job as we know it will soon be a thing of the past. This is one of the main arguments for a new system of social protection to replace the social security system and/or minimum wages: a Universal Basic Income (UBI). More ambitious than Speenhamland, this form of basic income would extend well beyond the working class as a universal citizen entitlement.
Some UBI advocates argue that the final nail in the coffin of traditional ‘welfare’ is the emergence of the ‘gig economy’ and widespread use of intelligent machines, which have been predicted to displace almost half of today’s jobs. This, they argue, would make room for an economy where we don’t need to work as long or hard as we do now. With fewer jobs, the link between social security and paid employment will become outdated and ‘work tests’ will be pointless.
In our search for the ‘shock of the new’, we sometimes forget that Australia has undergone three decades of labour market and social change, including expansion of the role of markets in a traditional domain of the State (for example VET and employment services), the re-casting of women’s roles in care and the formal labour market, the erosion of collective bargaining and the shrinking of industrial awards to a safety net, and growth in part-time and casual employment.
What’s changed and what has stayed the same?
Some of the above predictions are wide of the mark. It’s vital that we understand to what extent the old rules of the economy and welfare state have changed, and to what extent they remain the same.
First, and most obviously, public social expenditure has not collapsed despite decades of ‘austerity’ policies (see Figure 1). As quickly as old programs were cut, new needs were identified (especially in health care).
Figure 1: Public social expenditure has grown
Source: OECD Social expenditure data base
Rather than retrench social programs, governments have redesigned them and shifted priorities. Among the most important changes were ‘activation policies’, which prioritise employment for people of working age on social security payments, and the use of market mechanisms for government services.
One of the biggest shift in priorities in Australia and elsewhere was from benefits and services for people of working age (which have been neglected while their recipients are pilloried in the media), towards necessary improvements in retirement incomes and health care as the baby boomers retire (Daley & Coates 2016).
One sign of this bias against poorer, younger people is the failure in Australia to increase Youth and Newstart Allowances for unemployed people and students, whose real value has been frozen for the last 20 years. As a result, Newstart Allowance for a single adult is just $270 a week, $180 less than the pension. Pensions are indexed the way all social security payments should be: to wage movements, so that those on the lowest incomes share in the benefits of productivity improvements across the economy.
Another sign is the ‘welfare dependency’ mantra which blames unemployed people, sotto voce, for the cost of social security, even though unemployment payments comprise just 10% of overall Commonwealth social security and welfare spending (ACOSS 2017).
Over the past 25 years, expectations of paid workforce participation among social security recipients have intensified, under ‘activation’ policies pursed in most wealthy nations. This is part of an international trend towards higher workforce participation in wealthy nations, especially among women.
Activity requirements aren’t new for people we used to classify as ‘unemployed’: work tests have always been part of the eligibility conditions for unemployment benefits. What’s changed is the scope and intensity of these requirements (including supervised job search and joining training courses and other programs), and their extension to ‘new groups’ of working-age payment recipients. In 2003, the Howard Government launched an ‘Active Participation Model’ of employment services in which unemployed people had to follow a ‘service continuum’ of gradually intensifying requirements including Work for the Dole. Three years later these requirements were extended to many sole parents and people with disabilities under the ‘Welfare to Work’ policy.
Promotion of workforce participation, and help to secure a decent job, has potential benefits. Whiteford & Adema compared anti-child poverty policies across wealthy nations and found that a combination of adequate benefits and help to secure paid employment – as pursued in the Nordic countries – achieved the best results. Where many are excluded from paid work by structural changes in the economy, disabilities, or caring roles, public investment in paid work experience, training, child care and other employment-related assistance (including partnerships with employers to encourage them to take on people they would not otherwise employ) can make a difference.
Employment assistance is in need of improvement in Australia, where over two-thirds of people receiving Newstart Allowance have received income support for more than a year, often for the above reasons.
Not all countries have pursued the same ‘activation’ policies. It’s widely acknowledged that some countries aim to strengthen people’s work capacities and skills while others resort to ‘work first’ policies that impose activity requirements mainly to pressure people to leave benefits as quickly as possible. ‘Work for the Dole’ is a good example of the latter approach.
In Australia today, unemployed people are caught in a welfare trap but not the ‘intergenerational dependency’ the present government claims. They are caught between the high expectations and ever-increasing requirements of activation policies and the reluctance of governments and employers to invest in people who can’t simply walk into a job within a month or two of unemployment, even if the labour market picks up.
Our lean welfare state imposes maximum activation with minimum support. For example, unemployed people must generally search for 10 jobs a fortnight but consultant caseloads in ‘jobactive’ services of up to 300 have been reported. Many don’t receive the help they need.
As long as our two-tier social security system remains, ‘welfare to work’ policies mean that a growing number of parents and people with disabilities are shifted from pensions to the lower Newstart payment.
In recent years, the government has imposed new expectations on unemployed people that extend well beyond what can reasonably be expected to improve their job prospects: 15 to 25 hours a week of ‘Work for the Dole’, income management, and lately drug tests – all on the assumption that unemployment is caused by some kind of moral failure.
It seems the poor law is still with us.
Yet, activation policies challenge at least part of the old poor-law logic: if all who are able to join the paid workforce are required and supported to do so, there is no longer any justification to pay them less than those who cannot. Activation policies make the pension-allowance divide redundant, opening up the possibility of a ‘basic income’ that’s based on need rather than ‘deservingness’.
What’s changed in the Labour market?
Economists have been predicting the emergence of a ‘leisure society’ for a long time now. John Maynard Keynes famously predicted that by 2030 most jobs would be part-time and that ‘we shall do more things for ourselves than is usual with the rich today, only too glad to have small duties and tasks and routines.’ As John Quiggin points out, Keynes missed a few important trends including the conversion of much of traditional ‘women’s work’ into paid jobs. And of course, capitalism and modern marketing created new ‘needs’ which kept our noses to the (paid work) grindstone.
This is not the place to follow these debates in detail. The following relies much on an excellent paper by Borland & Coelli (2017), in which they track the impact of technological change on the Australian labour market over the last two decades, and assess whether jobs have become more scarce or precarious. Below are some of their key conclusions.
- Jobs and working hours have not declined overall
Since much growth in employment has been part-time, the volume of working hours per capita is a better summary of the state of employment than the share of people employed.
Figure 2 shows that, apart from the 70s and early 80s (which economist Bob Gregory called the ‘disappointing decades’), overall working hours increased from the 1960s to the Global Financial Crisis in 2008.
Figure 2: Hours of work per capita, Australia, 1965/66 to 2015/16, Actual hours worked series (Equals 100 in 1965/66)
Source: Borland J & Coelli M (2017)
- Robots won’t simply replace jobs, they will redistribute and redesign them
In 2013, Frey & Osborne made the explosive prediction that over the next decade or two from 2010, 47% of US jobs were at high risk from the spread of intelligent machines (computers and robots).
Others argued that this risk was unlikely to be realised in most cases, as intelligent machines complement the work of humans as well as displacing it. Borland & Coelli (2017)suggest that 10% or less of current jobs in Australia will go over the next few decades.
Whether we end up with the same number of jobs and hours per capita or not, there is a broad consensus that the labour market will continue to polarise between skilled and routine jobs (Figure 3), and manual labour and services (with the former gaining ground at the expense of the latter in each case):
Computerization has substituted for low-skill workers in performing routine tasks while complementing the abstract, creative, problem-solving, and coordination tasks performed by highly-educated workers.
As the declining price of computer technology has driven down the wages paid to routine tasks, low-skill workers have reallocated their labor supply to service occupations, which are difficult to automate. (Autor D & Dorn D (2013))
Figure 3: Share of employment by skill (Australia, 1986-2006)
Source: Borland & Coelli (2017)
So far, the losers have been workers with skills suited to routine jobs, especially routine manual jobs.
- Jobs are less secure for some, much the same for others
Another common claim is that jobs are less secure and more people are unable to obtain the working hours they need for a decent income. The real story here is mixed – with winners and losers depending on age and gender.
Figure 4 shows that a declining share of employees is leaving or losing their job each year, and that job turnover among women is converging downwards towards male turnover rates (which are gradually rising). Men still fare ‘better’ (to the extent that staying in the same job is a good outcome, which is not always so), but less so than in the past.
Figure 4: Rate of outflow from employment: Australia (1980-2016)
Source: Borland & Coelli (2017)
Figure 5 shows trends in un- and under-employment. That is the share of people who lack the paid hours they seek.
The striking trend since the early 1990s is the divergent experiences of young, middle-aged and older people. It is clear that young people have borne the brunt of growth in un- and under-employment since then. However, if mature age workers lose their old ‘steady job’, they seem to face a labour market just as hostile as that confronting young people.
Figure 5: Labour under-utilisation (unemployment + under-employment) by age (Australia, 1978-2014)
Source: Borland & Coelli (2017)
- The ‘gig economy’ is not new
Growth in precarious employment is not inevitable. Jim Sandford points out that for the first half of the 20th century a large share of jobs, especially low-skilled ones – were either casual (daily hire) or paid at ‘piece rates’ (according by work done rather than by the hour).
In the 1940s, wharfies lined up outside the docks each day to be told whether or not they would be given a job. ‘Uberisation’ is a high-tech version of the ‘hungry mile’.
The 1940s: The Hungry Mile
‘It was to this mile of wharves that maritime labourers in the nineteenth century and on into the 1940s, tramped each day regardless of the weather to find casual, low paid work, because that was the nature of waterfront work in those days.’ (Source: Rowan Cahill, Union Songs)
2010s: ‘The gig economy’
(Source: New York Times)
We managed to turn the tide against casualisation under conditions of full employment after World War Two. Information technology increases the risk, but also makes it easier to organise against it (for example, through union recruitment campaigns, cooperatives, and consumer boycotts).
There is a danger that if we concede defeat in the face of the ‘gig economy’, casualisation could become a self-fulfilling prophecy for a new generation of workers.
The threat to traditional forms of social protection is real, but it does not take the form that is popularly assumed. Jobs are not disappearing but workers with ‘routine skills’, and people entering or re-entering paid work after a long absence (including many young people, women and retrenched mature-aged workers) face greater risks of unemployment and job insecurity.
The welfare state is not shrinking but it is increasingly subjected to market disciplines, for example in the treatment of unemployed people.
In Parts 3 and 4 of this series, we ask whether a UBI or other variant of a basic income could help mitigate these risks.
This four part series is written based on the presentation, ‘From basic income to poor law and back again: can a UBI break the Gordian Knot between social security and waged labour?’, by Peter Davidson at the Australian Social Policy Conference at UNSW on 27 September 2017.
This four-part series explores the genesis of the idea of a ‘basic income’, how this evolved into a more broadly-based strategy for social improvement, the risks to job security and the welfare state, and the role of a basic income in overcoming them.
It featured recently in the Australian Tax Transfer Institute’s policy blog
Part 1 examines the surprising origins of basic income.
From Basic Income to Poor Law and back again
The first Basic Income scheme was introduced in Speenhamland, Berkshire in 1795, when the Napoleonic wars were underway, the French revolution was fresh in the minds of England’s rulers, and the industrial revolution was beginning.
Now that the idea has gained currency again, there is much to learn from Speenhamland and the Poor Law reform that followed it. The key lesson is that under capitalism, the labour market, politics and the welfare state are intertwined: changes in one impact on the other.
“That it is not expedient for the Magistrates to grant that assistance by regulating the Wages of Day Labourers, according to the directions of the Statutes of the 5th Elizabeth and 1st James: But the Magistrates very earnestly recommend to the Farmers to increase the pay of their Labourers in proportion to the present price of provisions;
“The Magistrates now present make the following calculations and allowances for relief of all poor and industrious men and their families, who to the satisfaction of the justices of their Parish, shall endeavour (as far as they can) for their own support and maintenance.
When the Gallon Loaf of Second Flour, Weighing 8lb. 11ozs. shall cost 1s, then:
every poor and industrious man shall have for his own support 3s. weekly, either produced by his own or his family’s labour, or an allowance from the poor rates, and for the support of his wife and every other of his family, Is. 6d.”
The Speenhamland system
We can see from this declaration in a pub in Berkshire that the idea of a Basic Income is not new. A cash benefit to meet basic living costs was paid out of council rates to thousands of farm workers (whether employed or unemployed). It was not universal (land-owners were not included), but this payment was widespread in the south of England.
Living standards of farm labourers in south of England were under pressure from high inflation driven by war, and a progressive loss of income from home production as new factories came into production across the north.
We can see from this proclamation that the local landlords (magistrates) were reluctant to impose minimum wages. Instead, they decided to use council rate revenues to protect the incomes of their workers.
The payments resembled modern income support, and even had their own equivalence scale to reflect the needs of families.
There was more. The landlords experimented with labour market programs for unemployed workers including subsidised private labour (‘roundsmen’), work for benefits (the ‘labour rate’), and (to a lesser degree) waged employment on public works
All of these anti-poverty policies are familiar to us today.
It is no accident this first ‘basic income guarantee’ coincided with the onset of the industrial revolution. This was one last push by landowners of southern England to keep their rural workforce in the face of industrialisation.
Apart from Tory noblesse oblige, this was about preserving the old system of rural labour relations in the face of emergent capitalism.
New wine in old bottles: income protection in England in the late 18th century
1. Minimum guaranteed income: The Speenhamland bread scale that provided specific amounts of aid in support of wages depending on the price of bread and the size of the family.
2. Seasonal unemployment insurance: During the winter months when agricultural work was scarce, some parishes provided unemployed farm workers and their families with a weekly stipend that varied depending upon family size.
3. Public works: Some parishes put the unemployed to work building roads or performing other types of work. Sometimes the supervision was done by public authorities and sometimes by private contractors.
4. Employer subsidies: Some parishes used poor relief funds to reimburse farmers and other employers who hired unemployed people. This was often called the ‘roundsman’ system because the unemployed workers would make the rounds of local employers.
5. Workfare: Some parishes allocated a certain proportion of unemployed people to each local employer with the idea that they would provide employment (at poor relief rates) instead of paying taxes for poor relief. This was referred to as the ‘labour rate’ system.
6. Child allowances: Many agricultural parishes provided a supplement to the income of male agricultural workers who had more than two or three children who were not yet of working age.
7. Workhouse: A minority of parishes required that unemployed people seeking relief enter a residential facility that imposed work requirements. Some of these facilities were publicly administered, and some were run by private contractors.
But it didn’t last long
When the industrialists gained power in Parliament in the 1830s, they argued that the Speehamland system depressed wages and encouraged idleness.
These views were supported by new ‘dismal science’ of political economy (especially by Ricardo and Malthus). Together with the industrialists they argued for a ‘free market’ in labour (with workers unable to organise or vote!). This was the stimulus for infamous 1830s ‘Poor Law inquiry’.
Their arguments are strikingly similar to those against a Basic Income today, and we have reason to be sceptical:
‘In sum, the Speenhamland myth was created in the years of agricultural downturn to divert blame for a deep agricultural crisis away from government policy and toward the rural poor who were the major victims of the economic downturn.
‘Many of the specific complaints in the historical record about the corrosive effects of the [Speenhamland] actually centre on ‘roundsmen’ or others who were engaged in ’make work’ activities.
‘When public agencies create employment specifically with the goal of making recipients work in exchange for relief, supervisors usually find it difficult to elicit high levels of work effort because recipients know that they are not working in a real job.’
‘Since the decision taken by the government on Ricardo’s advice to restore the prewar parity of the pound intensified the rural depression, the mythology worked to cover up the first major policy failure of the new science of political economy.’
‘By shifting the blame for the problems on to Speenhamland and all its pernicious evils, the economic liberals successfully reframed the agricultural downturn into a problem of individual morality and an enduring parable of the dangers of government ‘interference’ with the market.’ (Block & Summers 2003)
The new poor laws: a ‘stoic determination to renounce human solidarity’
The result was a national ban on public relief for able-bodied individuals outside the horrors of the ‘workhouse’. This was the birth of social security principles we know today: ‘less eligibility’ (that benefits should always be much less than minimum wages), ‘deserving and undeserving poor’ (a duty to work for the able-bodied or ‘undeserving’, and higher benefits for ‘the deserving’ or those ‘unable to work’ due to a disability).
As Polanyi put it in his influential history of the industrial revolution:
‘It was at the behest of these [1832 Poor laws] that compassion was removed from the hearts, and a stoic determination to renounce human solidarity in the name of the greatest happiness for the greatest number gained the dignity of a secular religion.’
‘The abolition of Speenhamland was the true birthday of the modern working class, whose immediate self-interest destined them to become the protectors of society against the intrinsic dangers of the machine civilisation. But whatever the future held for them, working class and market economy appeared in history together. The hatred of public relief, the distrust of state action, the insistence of respectability and self reliance, remained for generations characteristics of the British workers.’ (Polanyi 1954, p102)
Charles Dickens chronicled the horrors of the 19th century workhouse:
‘At present, if a boy should feel a strong impulse upon him to learn the art of going aloft, he could only gratify it, I presume, as the men and women paupers gratify their aspirations after better board and lodging, by smashing as many workhouse windows as possible, and being promoted to prison.’
The take-home message of the new poor laws was that in the nascent capitalist system, decent minimum incomes – the ‘social minimum’ – would not be guaranteed by public relief alone, they must also be underpinned by decent jobs, skills, and wages.
From Speenhamland to Nixon
The Speenhamland system was of more than academic interest to modern policy makers. When Richard Nixon revived the idea of a basic income in his ‘Family Assistance Plan’, he was warned against it.
In the Nixon Administration, Daniel Moynihan was tasked with developing a ‘Family Assistance Plan’. As Moynihan recalled:
“In mid-April Martin Anderson, of [Arthur] Burns’s staff, prepared ‘A Short History of a Family Security System’ in the form of excerpts on the history of the Speenhamland system, the late eighteenth-century British scheme of poor relief taken from Karl Polanyi’s ‘The Great Transformation’.
“The gist of Anderson’s memo was that in that earlier historical case, the intended floor under the income of poor families actually operated as a ceiling on earned income with the consequence that the poor were further immiserated.” (Block & Summers 2003)
What happened next: two roads to betterment
By the end of the 19th century, the labour movement and social reformers realised they would have work on two fronts to end poverty and deprivation: the labour market and the State, unions and the vote.
In Australia, by the end of the Second World War, these groups were successful in constructing the two pillars of the modern ‘social minimum’: labour market regulation and the welfare state.
In the labour market, this comprised:
- A high minimum hourly wage
- Full employment & regular working hours (for men)
- A high unionisation rate
In the welfare state it included:
- Free public education
- (Mostly) free public health services
- A robust public Vocational Education and Training system
- Age pensions
- An unemployment benefit safety net linked to a public employment service
- Family payments to prevent child poverty (both in & out of paid work)
It is often argued today that these social protections (at least for people of working age) can no longer be sustained in their present form; specifically that the social security system should now be replaced or supplemented by a Universal Basic Income. We explore the genesis of these arguments in Part 2 of this series.
This four part series is written based on a presentation on Basic Income by Peter Davidson at the Australian Social Policy Conference at UNSW on 27/9/17.
Whenever the media turn their attention to the Disability Support Pension (DSP), the unspoken assumption is that many of the recipients could have a job instead – if only the personal or political will was there. Recently-published research from the Melbourne Institute shows how challenging the labour market is for people with disability. It raises serious questions about the last decade of ‘welfare to work’ policies which have shifted people with disabilities from higher to lower payments while doing little to improve their job prospects.
The DSP ‘problem’
At 5.4%, the proportion of people of working age receiving the DSP is around the middle of the wealthy OECD countries.
Proportion of people of working age on disability benefits (2008)
Source: OECD (2010) ‘Sickness disability and work.’
Growth in reliance on the DSP has slowing for a decade, apart from a spike just after the economic downturn in 2008.
Source: Peter Whiteford, Crawford School ANU, April 2015
You wouldn’t know it from media reports that paint a picture of the relentless rise in the DSP, which is supposed to be caused either by ‘rorting’ of claims or a government strategy to conceal unemployment, depending on your ideological viewpoint.
The Welfare to Work ‘solution’
For a decade now, Governments have pursued ‘welfare to work’ policies to deal with the so-called ‘DSP problem’. These usually involve diverting people with disabilities from this pension payment to the Newstart Allowance, which is $170 a week lower. Similar policies have been pursued with sole parents, large numbers of whom were diverted from Parenting payment to Newstart. Recent analysis from the National Welfare Rights Network indicates that 173,000 people with disabilities – around one quarter of the total number of recipients – relied on Newstart Allowance in 2014. This is up from 96,000 in 2009 (an 80% rise in five years).
This has been handy for governments searching for budget savings, but it’s not a true measure of a genuine ‘welfare to work’ or ’employment participation’ policy. The true measure is not how many people are shuffled between payments – it is the proportion of people who are able to support themselves by moving into employment.
On that score, welfare to work policies for people with disabilities have failed. A recent report by Broadway and McVicarpublished by the Melbourne Institute concludes that the policy introduced by the Howard Government in 2006 to shift people with disabilities to Newstart and expand work requirements and supports had:
“no impact on the probability of being on welfare 12 or 24 months later” (p22)
The effects of welfare to work on the incomes of people with disabilities
We have known that the 2006 welfare to work policy had minimal impact on the job prospects of people with disabilities since its official evaluation was quietly released in 2009 (about a year after it was written). In my paper for the 2011 ASPC conference I used its estimates of the employment impact of the policy to calculate what proportion of those affected – people with disability newly applying for income support after 2006 – were financially better or worse off. They would be better off to the extent that the policy substantially improved their job prospects, and worse off to the degree that they were merely diverted to the lower Newstart Allowance without finding work.
This was one of many hotly contested issues at the time the ‘welfare to work ‘ legislation was brought before the Parliament. The Employment Department’s submission to a Senate Inquiry into the Bill asserted that:
‘Those people will absolutely be better off than if they were sitting on the disability support pension or parenting payment single without work’. (Dept of Employment and Workplace relations submission to Senate Community Affairs Committee Inquiry into the ‘Employment and Workplace Relations (Welfare to Work) Bill 2005′).
My conclusion was that only 3% of the first cohort of 97,000 people with disabilities affected by the policy during 2006-07 were better off (having secured fulltime jobs they would not otherwise have obtained) but 29% were worse off (having been diverted from the DSP to Newstart Allowance without finding fulltime employment). Two thirds were not affected by the change, as they would have been granted the DSP (or not) with or without the new policy.
Impact of Welfare to Work on the first cohort of people with disabilities affected (in 2006-07)
The mechanics of the 2006 Welfare to Work policy
The main components of the 2006 welfare to work policy were:
- Diversion of new applicants for income support assessed as having a ‘partial capacity to work’ (able to work 15-30 hours a week) from DSP (which they would previously have received) to NSA and other payments (which were usually much lower)
- New work requirements for this group, to search for and accept employment of up to 15 hours a week (under the DSP there were no activity requirements)
- Expansion of Job Network and Disability Employment Services to assist them to find employment (most went into the less-resourced Job Network rather than specialist disability employment services).
The idea was that, via a combination of new incentives (‘activation’) and supports (employment services), more people with disabilities would avoid long-term reliance on the DSP. There can be no doubt this was a high risk for people with disabilities at that time (and now). DSP recipients were more likely to die on the payment than to leave it for a fulltime job. So leaving the system as it is is not an option that will improve the lives of people with disabilities.
‘Activation’ policies can be effective in improving the job prospects of people on income support. The trouble was that the government mis-diagnosed the problem by putting more weight on individual incentives and less on employment capacity building and reform of the labour market.
What the official evaluation found
The official evaluation of the ‘welfare to work’ policy estimated that it increased the probability that people with disabilities who claimed income support during 2006-07 left income support (usually for employment) by 10 percentage points six months later. That’s not a bad result for any employment program, but the sting in the tail is that even after the reforms only 12% left income support within a year of claiming. In other words, their probability of leaving income support if they had continued to be granted the DSP without any activity requirements and supports was just 2%!
I could no longer find this evaluation on the Department of Employment website, but here is the graph comparing the probability of leaving benefits pre-reform (in 2015-06) and post-reform (2006-07). By comparing exits from benefits among these two cohorts of people with a ‘partial work capacity’, the evaluation estimated the impact of the policy. The impact is represented by the gap between the top two continuous lines in the graph.
Graph 9: Proportion of people with a partial work capacity who left income support before and after the ‘Welfare to Work’ policy
Source: Welfare to Work evaluation, p51
Note: Top continuous line shows the % who left income support after claiming during 2006-07 (after Welfare to Work). Green line near the bottom shows % who left income support after claiming in 2005-06 (before Welfare to Work). The ‘gap’ between the two lines is approx 10 percentage points after six months (26 weeks) on income support.
A ‘glass half full’ interpretation of these figures would conclude that: at least 10% more people with disabilities moved into employment as a result of the policy. But it isn’t necessary to reduce people’s income support by $170 a week to do this. It was the ‘activation’ and employment supports that did this, not the diversion of people from one payment to another.
What the Melbourne Institute report found
Even so, the Melbourne Institute research shows that the ‘glass half full’ conclusion was wrong. To assess the impact of a policy on reliance on income support, we also have to take account of the proportion of former recipients who return to benefits. That’s what the Broadway and McVicar study did. The table below shows their key conclusions regarding the impact of the welfare to work policy on reliance on income support 12 months after the 2006-07 cohort of claimants received income support.
Source: Data are from Broadway & McVicar, p18.
As the official evaluation found, the post-welfare to work exits from income support were much higher (37% instead of 24%) than before the policy was introduced. But the probability that people who left income support returned within the same 12 month period was also much greater (24% instead of 9%). This fully negated the benefits of the higher exits, so that in net terms the policy had no impact on the probability that people with disability had left income support 12 months after claiming it.
Further analysis revealed the main reason for the higher ‘return’ rate: the vast majority of new claimants for income support who were diverted from DSP to NSA subsequently successfully claimed the DSP. So in most cases the policy did not prevent access to the DSP, it postponed it. Nevertheless, a growing minority of people with a partial work capacity must have remained on NSA for a long time because we know the numbers of Newstart recipients with partial work capacity has ballooned since the policy was introduced in 2006. Few of them seem to leave income support altogether.
The elephant in the room: reform of employment practices
So why did welfare to work fail people with disabilities and what can be done about it? Broadway and McVicar nail it in their conclusion:
“Disability reforms need to do more than simply reduce the generosity and tighten the conditionality of payments – with the risk of exacerbating the already high levels of poverty experienced by people with disability in some countries – if they are to substantially impact on welfare dependency among people with disability. People with disability face barriers to employment that these measures, in isolation, are unlikely to overcome.One aspect missing from these reforms is any serious attempt to address the incentives of employers and potential employers of people with disability.” Broadway & McVicar (2015), p22.
These words should be chiselled in stone at the entrance of Parliament House in Canberra so that the next time Members and Senators vote on welfare reforms affecting people with disabilities they don’t back another policy failure.
Just desserts? Social security and ‘deservingness’
Who ‘deserves’ income support and why? An important new study on the social legitimacy of benefits sheds light on why some groups and some benefits are more favoured than others. This helps explain why the Newstart Allowance is $160pw less than the pension, why there are periodic ‘moral panics’ over the so-called ‘rorting’ of the DSP, and why people are concerned when millionaires receive Age Pensions. It also helps inform the design of social security payments.
It’s well known in Australia that some social security payments receive more support from policy makers and the public than others. For example, people receiving Age Pensions attract more support and respect than those relying on Newstart Allowance. Also, some groups of recipients (such as older unemployed people) are looked upon more favourably than others (such as younger unemployed people).
A new working paper from Wim van Oorschot and Femke Roosma from the University of Antwerp called ‘The social legitimacy of differently targeted benefits‘ explains why. It provides a useful framework for predicting which benefits and groups who attract the most (least) support. These issues are always at the back of the minds of social security officials as they design and reform benefits, and front of mind among politicians and the media. It’s time we understood them better.
Who is ‘deserving’
The paper begins by gathering the research evidence on public support for different social security payments and social groups in Europe and the USA (but the arguments are very relevant to Australia). It starts with common stereotypes about groups considered ‘undeserving’ of benefits: the ‘undeserving poor’, ‘lazy unemployed’, ‘black welfare queens’ (in the US), and migrants (especially in Europe). It then divides social security payments into those deemed more legitimate (pensions for retired people, payments that extend universally, payments for people with disabilities) and those deemed less so (payments for unemployed people, payments for single parents – especially in the US, and heavily means-tested payments).
All familiar stuff! It gets more interesting when they look at the history of European benefits. In virtually all wealthy nations, including Australia, the first social security payments were the ones that are still regarded as the most legitimate today:
‘from the end of the 19th century onwards first the schemes for the commonly most deserving categories of old, sick and disabled people were introduced, than family benefits and unemployment compensation, and lastly (if at all) social assistance for the least deserving category of poor people.’
Another revelation – religion played a role:
“social assistance arrangements in the Catholic European countries differ from those in Calvinistic countries based on different perceptions of the poor: in the Catholic perspective the poor are more seen in a traditional Christian way as a pitiful and deserving group of ‘children of God’, while in the Calvinistic perspective poverty is associated with the laziness and immorality of an irresponsible and therefore undeserving ‘underclass’.”
How benefit systems strengthen (and weaken) their own legitimacy
What they’re saying is that notions of deservingness were inscribed into the structure of social security payments from the beginning, and remain with us today. In Australia, the original Age Pension legislation required recipients to be ‘of good character and deserving of a pension’ . In the US, pensions for retired people are called ‘social security’ while payments for low-income sole parents are called ‘welfare’. The implication is that benefit categories can reinforce the stereotyping and stigma faced by many recipients.
Most wealthy countries have social insurance benefits where employees, employers and governments contribute jointly to social security funds to cover social risks such as old age, disability and unemployment. These contributions lend legitimacy to payments, since the recipients at one stage contributed to their cost – in accord with the principle of ‘reciprocity‘.
The trouble with these systems is that they are costly (since payments are often based on previous wages and are usually not means tested) and people unfortunate enough to lack a recent employment history (such as many unemployed young people, recently separated sole parents, and people with disabilities) miss out.
Almost uniquely, Australian social security payments are funded out of general public revenue. That means they are less legitimate in the eyes of many, but the system is more cost efficient in preventing poverty, and (in theory) noone in financial need is left out.
Yet the Australian system still classifies people according to a well-known hierarchy of deservingness:
- At the top are ‘pensions’ for older people, people with disabilities and caring roles which mean that people (supposedly) can’t undertake paid work.
- In the middle are family payments, traditionally paid regardless of the employment status of the parent.
- At the bottom are ‘allowance’ payments (formerly called ‘benefits’) for people who are deemed ‘able to work’.
Pensions are about $160 a week higher than allowances, despite the fact that financial disadvantage is more common among people on the lower allowance payments.
See the logic? Pensioners have a good ‘excuse’ for not supporting themselves financially through paid work. They are too old, or have disabilities or are caring for someone. Unemployed people don’t have a good excuse. This is the ‘control principle‘: payments attract less support the more we assume that recipients have control over their circumstances. It explains the ‘moral panic’ over the so-called ‘rorting’ of disability pensions by people who have less visible or easily classified conditions such as a mental illness or back pain. If they’re ‘able to work’, they don’t deserve a pension.
On the other hand, a recurring theme in Australian social security debate is our anxiety over ‘middle class welfare’. We have grown so used to having the most tightly targeted (means tested) social security system in the OECD that extending payments to people deemed well-enough off to look after themselves is frowned upon. Hence our concern about millionaires claiming Age pensions and families on $150,000 a year or more receiving family payments. This is the ‘need‘ principle.
Five tests of ‘deservingness’
In all, the paper identifies five dimensions of ‘deservingness’ that have an impact on social security policy (shown below): need, control, identity, attitude, and reciprocity.
This is of more than academic interest. If deservingness is inscribed into the payment structure as the diagram below suggests, then the benefit system itself may be reproducing and magnifying common stereotypes. Policy makers and the media quickly work out which payments are for ‘bludgers’ or the ‘undeserving or ungrateful poor’ and set out to make reliance on those payments ‘nasty, brutish and short’.
The pension – allowance divide
The historic distinction in Australia between pensions and allowances is based on deservingness. The idea behind this payment hierarchy is that people who ‘can’t work’ can be ‘looked after’ without encouraging long term reliance on benefits among those who can. The McClure Report’s proposal to introduce an intermediate stream of payments for people who ‘cant’ work right now but could with some help, including people with partial work capacity and parents caring for a school-age child, is a refinement of this idea.
Of course, policy makers claim economic and fiscal benefits for the present benefit hierarchy: that keeping payments low for unemployed people improves work incentives and reduces social security spending. The main problem with proposals for a universal ‘basic income’ that’s paid regardless of work capacity or job search effort is that people might not be encouraged to do what they can to support themselves.
As I explained in a previous blog, there are several problems with the pension-allowance hierarchy. An obvious one is that it undermines a core principle of our social security system that payment levels should be based on financial need.
Another serious challenge to the present payment structure has emerged in the last two decades. ‘Employment participation’ or ‘activation’ policies call into question the very basis on which people are classified into pension and allowances. Put simply, we no longer assume that people with disabilities, and parents of young children ‘can’t work’ (paid work, that is). Public policies over the last two decades have tried to remove barriers to employment for people who were previously excluded. This is a good thing.
Activation policies – requiring and supporting people on income support make the transition to employment – are a better solution to the so-called ‘free rider’ problem in social security than imposing poverty upon people. People who have at least some employment capacity can be helped through training, job placements, and other support to find stable employment. In return, it’s reasonable to expect that they take up those opportunities.
Trouble is, the pension-allowance distinction throws up hurdles to activation. If you have a disability and receive the DSP you stand to lose $160 a week once assessed as ‘able to work’ and transferred to Newstart Allowance. Activation strategies require governments to keep moving the goalposts between the two systems as more groups are ‘activated’. Sounds fine in theory but until the labour market adjusts to employing people with disabilities, these ‘welfare to work’ policies have left around 100,000 people on lower benefits with no improvement in their job prospects. The same goes for the 100,000 or sole sole parents shifted onto Newstart a few years ago, many of whom struggle to find paid work that is secure (not casual) and fits in with their caring responsibilities. Meanwhile they and their children have to live on a payment that’s at least $60 a week less than they would receive on the former ‘pension’ for single parents (now called Parenting Payment).
There’s another flaw in the system: two thirds of people on the lower Newstart Allowance are long-term unemployed. They have received this very low payment for more than a year. Not all of them have a disability or caring responsibilities. Many are simply the wrong age, have the wrong skills, or live in the wrong place to secure a job. And when the labour market is as ‘slack’ as it is now, even young, highly skilled workers face a high risk of prolonged unemployment.
Here’s where the categorical system falls down. We can’t invent a payment category to deal with every ‘legitimate’ reason that a person relies on income support for a long time. Nor can we accurately predict in advance which groups are most likely to do so. As activation policies are extended to more people who previously received pensions, the population on Newstart Allowance starts to look more and more like yesterday’s pensioners.
The explanatory power of ‘deservingness’
What policy insights can we glean from the research on perceptions of legitimacy of social security payments? It’s easy enough the say the present system’s unfair but the reality is that these perceptions of legitimacy drive much of the social security policy we’ve seen in recent years. Social security policy is shaped by stereotypes as much as rigorous analysis and research. These stereotypes are based on anecdotes, gossip, and often the family backgrounds and experiences of key Ministers. This wouldn’t be accepted in economic policy, but it’s par for the course in social policy and has been so for many years.
Why else has it proved so hard to raise the level of Newstart Allowance when every expert knows it’s way too low to meet basic living costs? Why did the previous Labor government review and increase pensions but not allowances? Why did the present Coalition government propose in last year’s Budget to remove income support altogether from young unemployed people for six months of every year? The answers are to be found in studies like this one rather than dry economic debates about the perennial trade-off between adequacy and work incentives (important as these are). Behind every work incentive argument lies a shadow debate over deservingness.
The paper cites research suggesting that perceptions of the legitimacy of benefits are deeply rooted in human psychology:
“in pre-historical, small-scaled societies we developed the skills to detect reciprocators who contribute to food sharing, and cheaters who violated the rules of cooperation by taking advantage of the collective gains. Dealing with these reciprocators and cheaters is assumed to have structured cognitive categories and created judgmental shortcuts, called ‘deservingness heuristics’ .Today we still apply these heuristics, which allows us to make more or less instant judgements about deservingness, also in relation to deservingness of social benefits. In these deservingness heuristics emotions play an important role.”
This doesn’t mean that social security policy is destined to be dominated by shock jocks and TV current affairs producers. The paper offers a way out:
‘deservingness heuristics can be ‘learned’ in the sense that being exposed to a certain ideology, culture of institutional setting leads to the development of certain deservingness logics.’
Tips for social security reform
The question this raises for social security policy – and it’s a topical one now that the Government is considering the proposals in the McClure Report – is ‘how can the benefit system be re-designed to improve fairness and workforce participation while at the same time reduce stereotyping and restore legitimacy?’ The stereotypes will always be there, but the social security system need not reinforce them in the way it does now.
Here are some suggestions:
1. Continue with the ‘activation’ strategy and strengthen it by investing in better employment assistance and training for people who have a longer road to travel to reach employment.
Activation is consistent with the ‘reciprocity principle‘: if it is implemented in a balanced and not harsh way it should strengthen the legitimacy of payments for unemployed people. It’s also a better way to maintain work incentives than keeping benefits too low to live on.
2. Remove the distinction between pensions and allowances for people of working age.
The base rate of payment for everyone of working age in need of income support should be the same. Additional costs faced by, for example, people with disabilities and sole parents, can be met using supplements that target those costs. This is consistent with the ‘need‘ principle.
It removes the ‘beauty contest’ over which groups ‘deserve’ a higher (or lower) level of payment. It also means that people can be ‘activated’ without bumping them down to a lower payment.
3. Retain population based categories to help determine what, if any, activity requirements should apply.
One potential problem with a standard base rate of payments from a legitimacy point of view is that the whole population of income support recipients might come to be viewed as ‘undeservedly poor’ (as are recipients of Municipal ‘welfare’ payments in many parts of Europe). Factors such as disabilities and caring roles do affect people’s ability to secure employment and this should be publicly acknowledged and taken into account. One way to do so is to retain payment categories such as disability, parenting and caring, and use them to decide what activity requirements (if any) should apply rather than levels of payment.
4. Keep family payments separate from income support payments for adults.
The purpose of family payments is to help parents with the costs of children. This purpose has more social legitimacy than income support for adults of working age, who might be able to support themselves (the ‘control’ principle). If blame is being apportioned for unemployment, children share no part of it.
Social security policy-makers of the 1980s and 90s were wise to separate family payments from income support, if only for this reason. Another strength of the family payment system is that it extends more broadly than a ‘welfare’ payment, to middle income families. This is widely accepted – provided they don’t extend to high-income families – since this is how the tax-transfer system takes account of the costs of children.
Extending family payments to middle income families also forges a bond between low and middle income families, consistent with the ‘identity‘ principle.
5. Be careful with language.
Call them social security payments, not ‘welfare’.
It’s time to end ‘welfare as we know it.’
Racing against time: indexation of social security payments
Indexation of social security payments is a dry subject. Does it really matter if they’re indexed to MTAWE or CPI? You bet! The Government’s budget proposals to index pensions only to the CPI and not to wages would reduce them by $80pw in a decade compared with present indexation, and save a motza. The Rudd Government’s decision in 2009 to index family payments to CPI instead if wages has saved over $1B and has already cost a low income family with 2 children under 13 years $19pw. If Newstart Allowance had been indexed to wages as well as the CPI over the last 20 years, it would now be $115pw higher ($52 instead of $36 a day).
How indexation works
Currently, pensions (including the Age Pension, Service Pension, Disability Support Pension and Carer Payment) are indexed twice each year by the greater of the movement in the Consumer Price Index (CPI) or the Pensioner and Beneficiary Living Cost Index (PBLCI). This is designed to keep pace with increases in the cost of living.
From time to time the base rate of the pension is also adjusted upwards to bring the maximum single rate up to a ‘benchmark’ of 27.7% of Male Total Average Weekly Earnings (MTAWE). This is designed to keep pace with community living standards. The Parenting Payment for sole parents is set at the lower level of 25% of MTAWE.
Looking forward: what will happen to pensions if they’re only indexed to the CPI?
Pensions play a vital role in preventing poverty. According to ACOSS’s ‘Poverty in Australia’ report, in 2012 13% of older people (over 64 years), 45% of sole parents on Parenting Payment, and 42% of people with disabilities on the DSP, lived below the poverty line. It would be a lot higher without the pensions to provide a safety net for those with no other source of income.
One of the most controversial policies announced in the 2014 Federal Budget was to index pensions to the CPI only from 2017, instead of both CPI and wages. The Government says pensions won’t be cut. Whether this is accurate depends whether this means means cutting the future spending power of the pension (in which case it’s not a cut) or cutting future payments compared with what people would have received under the present law (in which case it’s a cut). It sure looks like a cut to me – if not, then why does the Inter-generational Report estimate the Government would save over $40 billion a year (in current dollars) from this policy, together with the proposed increase in the pension age, in 40 years’ time?
We can see more clearly what impact indexing pensions to the CPI and not wages would have on future pensions from the following graph from the National Commission of Audit Report.
The maximum rate of the single pension under current policy (indexation to CPI and wages) is shown in the top (yellow) line. The second (blue) line shows the value of the pension if it is only indexed to the CPI up until 2027. The gap between the lines show how much pensioners would lose. In 2024 (a decade from 2014 when the Audit Report was prepared) the cut amounts to $80 per week.
Using estimates kindly provided by Peter Whiteford, if pensions are indexed to CPI only for the next 13 years as suggested in the IGR, they would fall from 28% to 24% of average earnings. If this continued over 40 years, it would fall to 16% of average earnings.
Not surprisingly, the Senate has other ideas. It has refused to pass this legislation, and the link between pensions and wages remains in place for now.
Looking back: What happened to pensions, NSA and family payments over 20 years?
The link between pensions and wage increases has been bipartisan policy for many years. The Whitlam Government set the single pension rate at 25% of MTAWE. The Howard Government put this into legislation in 1997. A decade later in 2007 the Rudd Government raised the benchmark to 27.7% of MTAWE (an increase above inflation of $32pw at the time). The 2014 Budget broke with that tradition.
To see what happens when payments are indexed differently, let’s look back at social security payment levels over the last 20 years from 1994 to 2014. To set the scene, this graph compares movements in the CPI and MTAWE over that period.
Average wages have risen strongly over the last two decades of solid economic growth, by around 2.5% above the annual inflation rate (CPI). This means that ‘real’ living standards of average wage earners have risen by about that amount each year. It’s obvious that if social security payments are indexed to the CPI only and not wages, their recipients will not share in future increases in the community living standards. Their living standards will be frozen in time. Wages are now growing more slowly, but in the long run we’d expect them to rise each year by about 1.5% to 2% above inflation.
The graph below shows rates of pensions, fortnightly Family Tax Benefit Part A fortnightly payments (FTBA) for 2 children under 13 years, and Newstart Allowance (NSA) since 1994. Rates are expressed in 2014 dollars (removing the effects of inflation). Only pensions have risen consistently above inflation – because they were indexed to wage movements and then increased further in 2009 (that would change if the 2014 Budget proposal become law). NSA flat-lined (apart from a small rise in 2000 to compensate for the GST). FTBA rose strongly in 2000 and flat-lined after that.
Gaps between pensions and other payments increased dramatically over this period. Living standards for pensioners rose (though the pension is still frugal) while those of low income families and unemployed people without any other source of income froze.
How did this happen?
Family Tax Benefit
Family Tax Benefit Part A (formerly Family Allowance) plays a vital role in staving off poverty among children. In 2012, according to ACOSS, 17% of children live below the poverty line. Most are in families reliant on social security and family payments, and about half are in sole parent families.
In 1988, the Hawke Government committed to reducing (actually, “ending”) child poverty. It proposed to do this by dramatically increasing family payments for the poorest families – both jobless and in low paid work.
The policy didn’t end child poverty but it did reduce it by one third: one of proudest social policy achievements of the Hawke Government.
As with pensions, this increase was made by linking (benchmarking) family payments to wages. So the maximum rate of family payments for a low income family could not be less than 16.6% of the pension rate for a married couple for each child under 13 or 21.6% for each child aged 13-15 years. Since pensions were linked to wages, family payments were also linked for the first time to wage movements. These benchmarks were also a bi-partisan commitment. They were retained by the Howard Government.
In 2000, family payments were increased substantially – above what was needed to compensate for the GST. As a result FTBA was above the benchmarks until 2008, and no further increases above inflation were awarded despite rises in pensions and wages from 2000 to 2008.
Then in 2009, the Rudd Government cut the link between Family Tax Benefit Part A and wages, so from then on it was only indexed to the CPI. This was announced at the same time as the increase in pensions, so this was a ‘saving’ (at the expense of low-income families) to pay for higher pensions. By 2013, the annual saving to the Budget was estimated to reach $500M.
Since sole parents were excluded from the 2009 pension increase (see above), this means they lost out in two ways. Their pensions were not increased, and their family payments are now lower than they would have been. Three years later the Gillard Government shifted many sole parents off the higher pension payment onto the lower Newstart Allowance – a typical payment loss of $60pw. At this time 25% of sole parent families lived below the poverty line.
So, just as low-income families were about to receive an increase in payments above inflation under the rules introduced by the Hawke Government and retained by the Howard Government, they were denied it by the Rudd Government. FTBA payments have not risen since in real terms, except under the Gillard Government’s ‘Clean Energy’ package to compensate for the carbon price.
By comparing movements in the CPI and MTAWE, we can work out how much low income families lost between 2009 and 2014 as a result of the removal of indexation of family payments to wages. A family with two children under 13 has lost around $19pw. It is inevitable that if family payments for low income families are not increased above the CPI, child poverty as it is usually measured (i.e. in relation to community living standards) will increase.
Unlike pensions and family payments, Newstart Allowance for unemployed people has never been indexed to wages, though the Whitlam Government did raise Unemployment Benefits to the pension level and the Hawke Government committed to restoring that link. Unemployed people have always been at the bottom of the social security pecking order.
Newstart Allowance is a good example of what happens to payments, and living standards, if they are only indexed to the CPI for a long time. It’s not a pretty sight. Living standards for recipients are frozen in time to 1994 (the last increase above inflation), when the internet was in its infancy. Newstart Allowance for a single adult is just $37 a day, nowhere near enough to meet basic living costs. If NSA had been indexed to average wages as well as the CPI, it would now be $115pw higher.
What should have happened
Now let’s see what would have happened to payment levels over the last two decades if NSA and family payments were indexed to wages consistently as pensions were, and if unemployed people received the same one-off increase granted to pensioners in 2009.
Newstart Allowance would have been $115pw higher and Family Tax Benefits for a family with two children under 13 would have been $19pw more (to show the impact of indexation to wages, the year 2000 increase in FTBA is removed – that payment would still be higher by 2014).
Not as simple as it sounds:
the Welfare Review’s social security reform proposals
The Welfare Review Report has finally been released. It advocates a “new flexible social support system that is simpler, sustainable, coherent and outcomes focused”.
The Review proposes to replace the present tangled mess (at left) of 20 income support payments paid at various levels (see figure at left) with ‘five main payments’:
- Tiered Working Age Payment
- Supported Living Pension
- Child and Youth Payment
- Carer Payment
- Age Pension.
Sounds good. But it’s not as simple as it sounds.
Most of the ’20 existing income support payments’ are actually sub-categories of three basic payments: pensions for people’ unable to work’, Allowances for people ‘able to work’ and student payments. The sub-categories relate to different population groups within each of the three payment types (e.g. sole parents, people with disabilities, unemployed people). These sub-categories play an important role: to identify limitations on people’s paid workforce participation (e.g. caring roles) in order to decide what range of activity requirements (if any) should apply. If they didn’t exist they would have to be invented.
The main flaw of the present payment structure
Most of the complexity and unfairness in the present system comes not from the population categories but from the division into the three different basic payment levels outlined above: pensions, allowances and student payments. The figure at left, from the ACOSS Welfare Review submission, shows that we have at least two distinct social security systems: pensions and allowances (student payments are included in the ‘Allowances’ box on the right), each with its own eligibility conditions, payment rates and income tests and activity requirements.
The old divide between pensions, allowances and student payments is based, not on financial need, but on outdated notions of ‘deservingness’ and ‘ability to work’. For example, people with disability are paid a pension because it’s assumed they won’t find paid employment and ‘deserve a pension’ – not because their financial needs are greater (which is a much better argument for higher payments for people with disabilities).
Over time, the gaps between pension and allowance payment rates have widened (see yellow bars in the graph at left). Unemployed people are paid the $36 a day Newstart Allowance because it’s assumed they might find a job tomorrow (even though the average time on that payment is 108 weeks!)
It turns out that the lowest payments go to the people in greatest financial hardship : Newstart Allowance and Parenting Payment, closely followed by DSP recipients (see graph at left).
If it’s bad for equity, the pension:allowance divide is also bad for work incentives. The closer people come to paid employment, the lower their payment. If I have a disability and am assessed as ‘able to work’, I drop from the DSP to the much lower Newstart Allowance. People on pension payments fear the loss of income security that dropping down to Newstart Allowance entails.
The Welfare Review’s proposed solution
The Review acknowledges these payment anomalies, which worsened in 2009 when pensions were increased but Newstart and student payments remained the same.
“People in similar circumstances, with similar basic living costs receive different levels of financial support and face different expectations of work.”
Yet when the Review’s proposals are unpacked, we find that at least one additional payment level would be introduced. The proposed ‘tiered working age payment’ has three different payment levels based on ‘ability to work’. So in place of three basic payment levels (three-and-a-half if we include the higher Newstart rate for sole parents) we would have four: the first tier, second tier and third tier of the new Working Age Payment (WAP), and pensions for carers and (some) people with disabilities. Is this simpler?
The Review side-steps proposals, advanced by ACOSS and others (and the original ‘McClure Report’ in 2001), that levels of payment (base rates and supplements such as a cost of disability supplement) should be determined on financial need rather than work capacity:
“A number of stakeholders suggested simplifying the payment system with a single base payment for people of working age, including all those with disability. This single base payment would be based on an assumption of a person’s current financial needs and not include consideration of capacity to work, now or in the future. A single payment system does not recognise the varying capacities of individuals to work. “
So instead of abandoning the old idea of different levels of payment based on work capacity (and whether people’s need for income support is ‘short term’ or ‘long term’), it proposes to refine it:
“There are some people who, because of severe disability, are not expected to work and as a result are likely to spend long periods of time on income support with no potential to supplement this support with earnings. This group will receive the Supported Living Pension.
For other people receiving income support, there is an expectation of work. A payment is needed that reflects the expectation that now or in the future people will be capable of some level of workforce participation. However, within this group, there are varying needs and capacities to work. Different tiers and rates are needed to reflect this diversity. These groups will receive the tiered Working Age Payment.”
The same argument is applied to parents and carers.
This is where the problems begin. By dividing people according to ‘work capacity’ the Review proposes that we should continue to ‘second guess’ people’s future work prospects and need for long-term income support.
For example in the proposed ‘tiered’ payment system people with disabilities would be divided (by the magic of work capacity assessments) into four groups receiving different levels of payment:
- unable to work more than 8 hours a week for the next five years
- unable to work more than 8-15 hours a week
- unable to work more than 15-29 hours a week
- able to work more than 29 hours a week
Is this feasible? Is it fair? The Report’s own data suggest that these days people spend a lot of time on so-called ‘short-term’ payments like Newstart Allowance even without disabilities or caring roles:
“The average income support duration for someone on Newstart with partial capacity to work is 277 weeks, while average Income Support duration of a person on Newstart Allowance, not assessed as having partial capacity to work, is 217 weeks. [NB: this includes continuous periods on other payments] “
And in the case of carers moving off the Carer Payment (a pension payment):
“Around 78.6 per cent of carers who exited from Carer Payment (before Age Pension age) in 2011 received further income support, such as Newstart Allowance, at some stage prior to March 2014”
It’s hard to escape the conclusion that the Review wants to fine-tune a flawed payment structure – to add a few new strings to a broken guitar and hope that it plays.
If the Review’s proposals were implemented, depending on your starting point in the present system, you could gain or lose. The graph below compares existing payment levels (first three bars) with the proposed system (next four bars) – note that the Review does not specify what the new payment levels should be.
- Some people with disabilities and sole parents on Newstart Allowance would move up to the ‘middle tier’ and GAIN (green bar at right)
- Other people with disabilities who apply for a pension under the new (more stringent) rules and miss out would go onto the ‘middle tier’ or ‘upper tier’ and LOSE (purple bar)
- It appears that adult students would move up from Austudy Payment to the ‘first tier’ and GAIN, provided the first ‘tier’ is equal to Newstart Allowance (green bar at left)
The proposed structure would lift some people off the lowest payments (and right a few injustices), but there’s no proposal to lift the lowest payment itself, despite acknowledgement that:
“During the consultations, many stakeholders raised concerns about the adequacy of allowances. In particular, stakeholders pointed to the increasing gap between pensions and allowances.”
The ACOSS proposal
Is there a better alternative?
In its submission to the Review, ACOSS proposed replacing pensions allowances and student payments for people of working age with a ‘base rate and supplements’ model of income support in which payment levels are based solely on financial need, not work capacity. Here’s what the proposed structure would look like (at left, also from the ACOSS submission).
It turns out that people with disabilities, sole parents and carers do need higher payments – not because of a supposed ‘inability to work’, but because they face higher costs (the costs of disability and care). Those costs would be explicitly acknowledged in the proposed system.
People with disabilities, for example, would receive the same base rate of payment as other social security recipients plus a supplement to assist with the costs of disability, which they would keep if they moved into a job paying a modest wage.
The graph at left illustrates how this ‘base rate and supplements’ model (with a costs of disability supplement) would affect social security payments for people with a ‘partial work capacity’ (the two bars at left) and those presently entitled to the DSP (the two bars at the right). Those on Newstart Allowance would GAIN (red bar at left) and those on DSP would receive the same level of payment (red bar at right), unless they obtain employment, in which case the Disability Supplement would leave them better off than under the present system, improving work incentives.
Work capacity assessments (and population based categories such as disability, carer, and parenting) would still be needed to determine whether activity requirements should apply (to give people a degree of certainty about the range of requirements that might apply to them if they have a disability or caring roles). But they would not determine payment levels.
Not a simple system, but much simpler and fairer, and better for work incentives, than the labyrinth we now call a social security system.