Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
The poverty of welfare reform discourse. Prepared for the National Social Policy Conference, University of New South Wales, Sydney, 4-6 July 2001.

Download PDFDownload PDF

Henman, P. (2002), ‘The poverty of welfare reform discourse’, in T. Eardley and B. Bradbury, eds, Competing Visions: Refereed Proceedings of the National Social Policy Conference 2001, SPRC Report 1/02, Social Policy Research Centre, University of New South Wales, Sydney, 180-191.

The Poverty of Welfare Reform Discourse

Paul Henman Department of Sociology, Macquarie University

1 Welfare Reform Discourses1

During the last decade, the need for welfare reform has been an often repeated refrain in the First World, but especially in the English-speaking countries of Australia, the United Kingdom, the United States of America, New Zealand and Canada. The various reasons for exhorting welfare reform reflect a wide political range of ideas about the economy and society.2

In Australia, the former Minister for Family and Community Services, Jocelyn Newman (1999), argued that there had been a large increase in welfare dependency that needed to be reduced. Alongside this, the Howard Government’s clear interest in extending the ‘mutual obligation’ principle in welfare presents a view that the previous welfare system lacked obligation. The Work for the Dole program also emphasises the acquisition of a work ethic that welfare recipients supposedly lack.

In response, the Reference Group on Welfare Reform (RGWR) (2000a and 2000c) championed the notion of social and economic participation as a way to interpret the need for welfare reform and to shape the form it might take. The language implies that welfare recipients do not participate in Australian society or the economy.

In response, and in its earlier social security and unemployment reforms, the Australian Labor Party has continually championed labour market training programs to address the perceived lack of skills of unemployed persons, a lack which is supposedly self-evident due to the very status of being unemployed (Prime Minister, 1994). Buttressing these programs has been the utilisation of a ‘reciprocal obligation’ rhetoric and advocacy of income tax credits to improve incentives.

The rational actor framework espoused by economists leads them to emphasise either the apparent lack of skills to enable the unemployed to compete in the labour market or a lack of incentives to take up paid work.

Australian Third Way advocates seem to have absorbed ideas from a range of viewpoints, often referring to a ‘paid inactivity’, a lack of reciprocity, ‘negative welfare’ and the passive welfare state (Latham, 2001a; 2001b; Pearson, 2001).

1 Although I personally prefer the term social security to welfare, to avoid confusion and to maintain a level of consistency with current terminology, this paper will use the latter term.

2 For a more comprehensive comparison of welfare reform that reflects the great range of welfare regimes see, for example, Esping-Andersen (1996) Giersch (1997) and Lødemel and Trickey (2001).



This diverse range of Australian welfare reform discourse is repeated in other countries. Lawrence Mead’s (1986; 1997) call for obligation has been influential in the USA; former American President Clinton sought to improve incentives to ‘make work pay’; British New Labour has also negotiated a discursive and policy framework involving obligation and labour market skills in its New Deal (see Grover and Stewart, 2000); and the OECD has for many years called for the creation of more active welfare state policies (2001). In a similar way, Anthony Giddens (1998), the academic Third Way advocate, argues that the welfare state has created moral hazard and welfare dependency and that it needs to be restructured to enhance obligations and investment in human capital.

2 Welfare Reform Analytics

Despite the diversity of these welfare reform discourses, they significantly share a similar analytical focus, a comparable ‘welfare reform analytics’. Each of these discourses begins by treating the behaviour of welfare recipients and the welfare recipient population as problematic. In each case, recipients lack something: obligation, initiative, skills, motivation, entrepreneurial zeal, incentives, work ethic, normalcy, and so on. The task is then to restructure welfare so that it provides a new regime of carrots and sticks which then changes the problematic behaviour.

As a consequence of this welfare reform discourse, the problem which welfare reform seeks to address is clearly identified as being located in individual welfare recipients, the recipient community and/or the welfare system that co-constructs welfare recipients and communities. The analytical gaze of the welfare reform analytics is accordingly blind to the structural nature of the welfare realities it seeks to understand and respond to and to the activities and behaviour of non-recipients3. In contrast, this paper shows that structural changes in Australia’s economy and society, not recipient behaviour nor the welfare system explain much of the increase in recipient numbers over the last 35 years. .

A further consequence of the welfare reform analytics is that the policy ‘solutions’ that are identified focus on changing welfare recipients, in their behaviour or skill levels. The efficacy of policy aims is apparently self-evident.4 In contrast, the questions of whether changing the structural realities would be more effective or whether changing recipient behaviour without addressing the structural realities will be ineffective are not even raised, let alone seriously considered.

As a consequence of the focus of the welfare reform analytics, policy change almost universally involves increasing the governmental requirements on recipients (or some section of them) with some approaches adding in a dose of government expenditure for good measure. Under the threat of their already inadequate level of income, recipients are obliged to be more active and jump more hurdles. The resulting policies

3 This is assuming that a dichotomy between recipients and non-recipients can be clearly established, which it cannot. For example, many recipients combine part-rate benefit with part-time wages, welfare is provided to low-income families as in-work benefits, and many others receive tax benefits in lieu of welfare benefits.

4 To be sure, the policy objectives of much welfare reform discourse are unclear. Although in some cases this probably reflects lack of intellectual clarity, in other instances the ambiguity is intentional as part of a political tactic.



thus reinforce the social and economic disparities between those in full-time paid work and those that are not.

3 The Structural Realities of Welfare

Despite the unerring focus of welfare reform discourse on problematising and changing the behaviour of welfare recipients, this paper demonstrates that the structural realities of social and economic change explain much of the increase in recipient numbers. This will be shown through an analysis of the change in recipient numbers since 1965.5

Figure 1: Proportion of Population of Workforce age Receiving Income Support Payments, 1965-1998

Source: Newman, 1999

5 This section of the paper draws heavily on Henman (2001).












3.7% 3.5% 3.2%









5.7% 5.9%












1965 1968 1971 1974 1977 1980 1983 1986 1989 1992 1995 1998

Proportion of workforce age people in receipt of Social Security payments Unemployment rates (August original ABS)



The implications of this finding are that policies that seek to change recipient behaviour without structural change are misplaced and will be marginally effective at best, and that improved labour market outcomes for recipients can only be obtained through cross-portfolio policy change involving multiple levels of government and actively engaging the community and business sectors.

Presenting its case for welfare reform, both the Government and the Reference Group on Welfare Reform repeatedly presented graphs depicting a considerable increase in recipient numbers since 1965 (see FaCS, 1999; Newman 1999; RGWR 2000b). Furthermore, levels of recipient numbers were often graphed alongside official unemployment or ‘jobless’ statistics to allegedly show that the increase in recipients is only partially due to increases in unemployment. (See, for example, Figure 1 reproduced from Newman, 1999). The apparent implication is that the discretionary behaviour of recipients to withdraw from the labour market accounts for the difference. Unfortunately, the arguments presented and conclusions reached by the Government and the Reference Group are fallacious.

Rather, as the Australian Bureau of Statistics has recently observed,

[t]his increase [in income support receipt by the workforce-age population] is largely associated with: sustained declines in full-time employment and increased levels of unemployment, particularly among young people and those just below retirement age; increasing proportions of people without partners in general, and lone parents in particular; and increasing levels of education participation among young people (ABS, 2001: 106)

Thus, in interpreting changes in the recipient population, a number of important factors need to be isolated and assessed. These are economic cycles, artefactual distortions resulting from policy change, labour market transformations and demographic changes.

Economic Cycles

It is well known that economic cycles affect aggregate levels of employment. The economic downturns of the late 1970s, early 1980s and early 1990s led to reductions in employment and increases in unemployment (see Figure 1). These cyclical increases are also evident in recipient numbers (see Figure 1). The dynamics of the economic cycle would suggest that unemployment numbers should return to pre-recession levels when growth returns. However, there has been a ratcheting-up of the ‘natural rate of unemployment’ which is largely related to structural changes in the labour market and economy.

Artefactual Distortions

The first problem with a straightforward examination of recipient numbers is that policy changes, including the increasing generosity and coverage of payments, alter recipient numbers without a change to recipient behaviour or social and economic realities. In interpreting changes in recipient statistics, these represent artefactual distortions. This is because it is policy change, rather than underlying changes in recipient behaviour or social and economic realities, that is being measured. Figure 2 illustrates changes in recipient statistics when the two largest distorting factors are



accounted for, namely, the need to include Service Pensions to war veterans, their wives and widows (Adjustment 1), and the removal of Parenting Allowance to low-income households (Adjustment 2). Service Pensions need to be included, as they have been to some extent a substitute for Disability Support Pension in the past, whereas recipients of Parenting Allowance for low-income households must be removed because the policy change is a form of in-work benefit and it involves an increase in benefit generosity.

Although these adjustments do not challenge the large historical upward trend in recipient numbers, they do significantly suggest that the rise in recipient numbers from the recession peak of the early 1980s to the peak in the mid 1990s is less pronounced than at first suggested. Furthermore, the adjusted data indicate that the measured rise in recipient numbers from 1995 solely results from the introduction of the in-work benefit of Parenting Allowance to low-income households. Once this distortion is removed, there has been a decline in recipients since 1995.

Figure 2: Income Support Recipients as a Percentage of the Workforce age Populations






1965 1970 1975 1980 1985 1990 1995 2000

% of workforce age population

Income Support Recipients Adjustment 1 Adjustment 2 Adjustment 3

Source : Author’s calculations from published statistics from the Department of Social Security and of Family and Community Services.

Labour Market Transformations

Since the 1960s, there have been significant transformations in the labour market. Part-time work, particularly for women has grown, while full-time employment especially in male manufacturing jobs has declined. Female participation in the labour market has dramatically increased, while the youth labour market has been decimated. Deregulation, downsizing and outsourcing have also increased the existence of precarious - that is, casual, part-time and contract - employment.

The enormous growth in part-time work has especially impacted on recipient numbers. Because the growth in part-time work has been accompanied by a loss in



full-time work, this has involved an increase in persons combining part-time wages with part-rate benefits, rather than a transfer of full-rate recipients to part-rate recipients. In short, growing part-time work has exacerbated recipient numbers, rather than providing avenues for people to partially move off benefits. Figure 2, Adjustment 3 indicates the effect of removing from the recipient population people who combine part-time work with part-rate benefits. Recognition of the existence of such people is extremely important when comparing recipient statistics with unemployment statistics as was done by the Government (see Figure 1). This is because a person who is employed for one hour a week or more is regarded as employed and not unemployed.6

A further problem concerning the comparison of recipient statistics with unemployment statistics is that the latter are not a good reflection of the number of people who would like work but are unable to find it. In particular, people who would like work but who are not currently looking for work (such as discouraged jobseekers) are not included in the statistics, nor are those who want work but were not able to begin work immediately (for instance, those who are sick, are full-time students or need to make prior work arrangements). The Australian Bureau of Statistics defines such people as having ‘marginal labour force attachment’.

For these reasons, unemployment statistics do not provide an accurate picture of the extent of structural joblessness - that is, the number of people who want work but have not been able to obtain it. Figure 3 shows that since 1977, when the first statistics were collected, the number of people marginally attached to the labour market has almost consistently been greater than the number of official unemployed persons, and has oscillated between six and eight per cent of the workforce age population. In addition, people who want work but who are not available to start work in the next four weeks are regarded as not in the labour force. These individuals have been separately counted by the ABS since 1983 (the dotted area)

Figure 3 compares the total number of jobless persons who want work with the number of full-rate income support recipients (reproduced from Adjustment 3, Figure 2). Part-rate recipients have been excluded from the comparison because as they have part-time work, they are officially regarded as employed. The comparison shows a remarkable level of concurrence between the number of jobless income support recipients and the number of jobless who want work.

These statistics, however, must be interpreted with caution. In particular, not all of the jobless who want work are welfare recipients. The former category includes, for example, people with partners in the labour force and full-time students. Counterbalancing this effect is the observation that many income support recipients do not currently want work, such as disabled adults or full-time carers of children or frail/disabled adults. Despite the need for caution when interpreting the statistics in Figure 3, they strongly support the view that many income support recipients want work, but are unable to find it. This means that structural factors in the labour market accounts for most of the increase in welfare receipt rather than lack of incentive to work or welfare dependency.

6 Australia’s unemployment statistics are based on a national survey on labour force participation and not on benefit numbers. The survey defines a person as unemployed if they are 15 years and over, currently available and looking for work and did not work for one hour or more in the week prior to being surveyed (ABS, 2000).



Demographic change

The foregoing analysis highlights the importance of structural unemployment and underemployment as primary reasons for the long-term increase in welfare recipients. However, this outcome is not simply a result from falling (male) full-time employment. Overall, labour force participation has increased over the last few decades, particularly by married women. Changes in recipient numbers and structural unemployment are thus outcomes from a complex set of changes in the labour market interacting with changes within households.

People tend to partner others with a similar educational level, socio-economic status or class. From after World War II to the mid 1970s, the breadwinner model was the dominant form of household organisation.7 The male was employed full-time while his female partner was responsible for maintaining the household and rearing their children. Because husbands were assumed to provide financial support for their wives, these households did not receive income support. (Schematically, see the left-hand side of Table 1.) Since the mid 1970s recession that ended full employment, the key changes in the labour market include a loss of low-skilled, low-paid ‘male’ jobs, especially in manufacturing, and an increase in high-skilled, high-paid jobs. This has resulted in a loss of jobs for men in low-skilled or working class households and an increase in employment by women, particularly in high-skilled households. Even if the number of (full-time) jobs had remained constant, the net result would be an increase in welfare recipients due to the loss of employment in low-skill households. In short, there has been a net transfer of jobs to high-skilled or middle class households to the cost of low-skilled, working class households, resulting in increased welfare numbers. (See right-hand side of Table 2.)

To further complicate matters, since the mid 1970s, there has been a shift from two-adult to single-adult households. This change is due to a number of reasons, including marriage and relationship breakdown and divorce, a greater social acceptance of sole parents, delayed partnering due to extended education and an increase in people deciding not to partner. The shift to single adult households by itself is likely to increase welfare receipt. This is because of the statistical reality that single adult households are more likely to have no adults in work - and therefore require welfare - than two adult households are. This is schematically represented by the dotted lines on the right-hand side of Table 1.

In summary, most of the increase in recipient numbers from the mid 1960s occurred during the period 1971-1983. The key reasons for this growth include reductions in low-skilled (male) full-time employment and the increase in single-adult households.

7 As Australia’s benefit system only recognises heterosexual relationships, same sex couples continue to be treated as independent individuals.


Figure 3: Full-rate Income Support Recipients and Jobless Who Want a Job as a Percentage of Workforce A

g e Po

p ulation






196 5







% of workforce age


Marginal labour force

Want work - not immediate start

Full-rate recipients

Source: Author's calculations from

p ublished social securit

y statistics and from ABS statistics

( Catalo

g ue Nos. 2203.0, 2220.0

) .



Table 1: Changes in Household Structure, Employment and Income Support Receipt

Pre mid-1970s Post mid-1970s

High skilled household

♂ full-time job

no income support

♂ full-time job

♀ unpaid domestic work

no income support ♀ full-time or part-time job

no or part income support

Low skilled household

♂ unemployed

1 income support payment

♂ full-time job

♀ unpaid domestic work

no income support ♀ unpaid domestic work

1 income support payment

To further complicate matters, since the mid 1970s, there has been a shift from two-adult to single-adult households. This change is due to a number of reasons, including marriage and relationship breakdown and divorce, a greater social acceptance of sole parents, delayed partnering due to extended education and an increase in people deciding not to partner. The shift to single adult households by itself is likely to increase welfare receipt. This is because of the statistical reality that single adult households are more likely to have no adults in work - and therefore require welfare - than two adult households are. This is schematically represented by the dotted lines on the right-hand side of Table 1.

In summary, most of the increase in recipient numbers from the mid 1960s occurred during the period 1971-1983. The key reasons for this growth include reductions in low-skilled (male) full-time employment and the increase in single-adult households.

4 Policy Implications

The identification of the structural changes outlined above is commonplace. However, their location in welfare reform discourse is varied. For example, former Minister Newman directly linked these structural changes to increasing welfare receipt:

as a result of these pressures [of labour market restructuring, increased labour market participation of women, population ageing and early retirement and the unequal distribution of work], the past 20 years has seen a large increase in the proportion of workforce-age people on social security payments (Newman, 1999).



In contrast, the Reference Group on Welfare Reform refers to such structural change to argue that the current welfare system is anachronistic (RGWR, 2000c), a view also evident in the Third Way literature.

Even when the enormous effect of such structural transformations on recipient numbers is recognised, welfare reform advocates proceed to espouse policy changes that completely ignore such transformations. Indeed, given that these transformations account for virtually the entire increase in welfare receipt, any attempt to reduce recipient numbers without responding to these transformations is simply playing around the edges and doomed to fail to produce any substantial outcome.

Without job growth, increasing the activity or skills of the unemployed, the obligations of welfare recipients and incentive regimes, will only exacerbate a sense of failure and worthlessness, the fragility of their income and divisiveness between the haves and the have-nots. The best that can be hoped for is a reduction of the time people are out of work by rotating a greater pool of people in and out of work.

A further illustration of the poverty of the welfare reform discourse is its failure to identify key policy initiatives that could be undertaken even within the reformers’ narrow focus. In particular, unemployment can occur when there is a skills mismatch - that is, a shortage of appropriately skilled persons to take available jobs.8 No Australian welfare reform document identifies the areas of skills mismatch nor analyses the extent of this problem. Accordingly, training policies can not be devised to address this problem, and so labour market programs are unable to maximise their impact. Similarly, geographical mismatch, whereby skilled unemployed persons are located in different areas from job vacancies that might use their skills, has not been analysed nor have remedial policies been identified. Welfare reform deliberations have not considered the ways in which employers may discriminate against the long-term unemployed, older workers, Indigenous persons or those with non-English speaking backgrounds, thereby creating barriers for people trying to enter or return to the labour force. Furthermore, welfare reform discourse fails to engage seriously in the question of the adequacy of welfare benefits and to recognise that lack of income can undermine people’s jobsearch and reskilling activities, their health and thus their capacity to rejoin the labour market. Regardless of the failure of welfare reform discourse to recognise that the structural realities are a primary policy concern, these shortcomings suggest that welfare reform has been stifled by a framework emphasising the inactivity and inability of the jobless and the passiveness and lack of incentives of existent welfare policy.

It is of course possible, that some advocate welfare reform not in an effort to reduce welfare expenditure or the poverty associated with joblessness, but as a political tactic. Wedge politics is alive and well in Australia (and elsewhere) and marginalising and demonising welfare recipients is likely to gain support from tabloid newspapers or from voters with downward envy (Gunn, 1996). To the extent that this is the case, then policy outcomes are irrelevant.

Assuming that policy, rather than political, outcomes are desirable, when identifying policy responses to the current state of welfare, it is first necessary to understand the nature of the policy problem to be addressed. In much of the welfare reform discourse, the significant increase in recipient numbers is taken as prima facie evidence that the welfare

8 Despite the focus on a lack of skills, there is some evidence to suggest many people are over-qualified for their jobs.



state is not working, that it is broken and in need of repair; that welfare has created moral hazards, fostered dependency and delivered rights without obligations. However, nothing can be further from the truth. As has been shown, it is the structural social and economic transformations that have created the increase in recipient numbers. Consequently, by financially supporting those who have been involuntarily excluded from the labour market, the welfare system is working and is far from broken. Put another way, the increase in recipients is the consequence of labour market policies of deregulation and downsizing, public service disinvestment and modest fiscal and monetary policies. Indeed, the safety net of the welfare state enables capitalism to become more laissez faire and to go global. As a consequence, the current crisis is not to do with an allegedly antiquated and inappropriate welfare state, but stems from the rise of poverty, inequality and social exclusion that has resulted from decades of structural change.

This analysis, therefore, highlights the importance of identifying policies that seek to alter the structural transformations to create greater opportunities for people to move from welfare to work. In contrast to welfare reformers’ concentration on the supply side, we need to give primary attention to the demand side of the labour market. This also means that the policy focus must widen from the isolated treatment of welfare policy to consider the interconnectedness of policies and institutions that constitute the nature of the labour market. Such policies must involve monetary and fiscal settings, taxation, industrial relations, employment, public services and infrastructure investment, research and development, regional development and education. It will necessarily involve what British New Labour calls ‘joined up government’. It will necessarily involve a critical rethinking of neo-liberal economic policies to ensure a greater mutuality of social and economic objectives. It also means that instead of accepting the globalisation rhetoric that says nation states are helpless to act, we actively engage in a politics that can shape economic globalisation to achieve more equitable and democratic outcomes Mishra, 1999; Kerr, 2001). Welfare reform discourse therefore needs to widen its horizons, as William Beveridge sought to do, to actively construct a social and economic environment that can realistically achieve the aims of welfare reform.

Fortunately, there are many well-qualified authors who have already identified and advocated policies that can create jobs and improve the conditions for the most disadvantaged, and are appropriate for today’s realities (for instance, Rees et al., 1993; Langmore, 1994; Morris, 1996; Galbraith, 1998; Stilwell, 2000). The task now is to see them to fruition.

References Australian Bureau of Statistics (2000), Labour Force, Australia, Catalogue No. 6203.0, ABS, Canberra. Australian Bureau of Statistics (2001), Australian Social Trends 2001, Catalogue No.

4102.0, ABS, Canberra. Australia, Department of Family and Community Services (1999), ‘Trends in pension and benefit receipt’, Research FaCS Sheet, No. 2, Canberra. Esping-Andersen, G., ed (1996), Welfare States in Transition: National Adaptions in

Global Economies, Sage, London. Galbraith, J. K. (1998), Created Unequal: The Crisis in American Pay, Free Press, New York. Giersch, H., ed (1997), Reforming the Welfare State, Springer, Berlin. Giddens, A. (1998), The Third Way: the Renewal of Social Democracy, Polity Press,




Grover, C. and J. Stewart (2000), ‘Modernising social security? Labour and its welfare-to-work strategy’, Social Policy and Administration, 34(3), 235-252. Gunn, M. (1996),‘Downward envy: why middle Australia hates the poor’, The Australian: Weekend Review, 2-3 November, 1-2. Henman, P. (2001), ‘Deconstructing welfare dependency: the case of Australian welfare

reform’, Radical Statistics, No. 79. Kerr, D. (2001), Elect the Ambassador!: Building Democracy in a Globalised World, Pluto Press, Sydney. Langmore, J. (1994), Work for All: Full Employment in the Nineties, Melbourne

University Press, Melbourne. Latham, M. (2001a), ‘Stakeholder welfare’, presented at the International Conference on Asset-Based Welfare, January, Institute of Public Policy Research and Centre for Social Development, London. Latham, M. (2001b), ‘Making welfare work’, in P. Botsman and M. Latham, eds, The

Enabling State: People Before Bureaucracy, Pluto Press, Sydney, 115-131. Lødemel, I. and H. Trickey (2001), ‘An Offer You Can’t Refuse’: Workfare in International Perspective, Policy Press, Bristol. Mead, L. (1986), Beyond Entitlement: the Social Obligations of Citizenship, Free Press,

New York.

Mead, L. ed (1997), The New Paternalism: Supervisory Approaches to Poverty, Brookings Institution Press, Washington, D.C. Mishra, R. (1999), Globalization and the Welfare State, Edward Elgar, Cheltenham. Morris, P. ed, (1996), Jobs and Justice: a Public Service Agenda for Full Employment,

Lawrence and Wishart, London. Newman, J. (1999), The Challenge of Welfare Dependency in the 21st Century, Discussion Paper provided to the Welfare Reform Reference Group, Department of Family and Community Services, Canberra. Organisation for Economic Cooperation and Development (2001), Labour Market Policies

and the Public Employment Service, OECD, Paris. Pearson, N. (2001), ‘Rebuilding Indigenous communities’ in P. Botsman and M. Latham, eds, (2001), The Enabling State: People Before Bureaucracy, Pluto Press, Sydney,


Prime Minister (1994) Working Nation: Policies and Programs, Australian Government Publishing Service, Canberra. Rees, S., G. Rodley, and F. Stilwell, eds, (1993), Beyond the Market: Alternatives to Economic Rationalism, Pluto Press, Sydney. Reference Group on Welfare Reform (2000a), Participation Support for a More Equitable

Society: Interim Report, Department of Family and Community Services, Canberra. Reference Group on Welfare Reform (2000b), Interim Report of the Reference Group on Welfare Reform: Technical and Other Appendices, Department of Family and Community Services, Canberra. Reference Group on Welfare Reform (2000c), Participation Support for a More Equitable

Society: Final Report, Department of Family and Community Services, Canberra. Stilwell, F. (2000), Changing Track: A New Political Economic Direction for Australia, Pluto Press, Sydney.