Room to move? Professional discretion at the frontline of welfare-to-work.

McDonald, Catherine ; Marston, Greg

Illuminating the day-to-day operations of policy allows decision makers and researchers to engage in candid evaluations of the developing welfare-to-work policy model and the politics and principles that inform it. We focus on policy implementation from the perspective of front-line workers--in this case social workers employed by Centrelink. Specifically we examine the conditions of exercising discretion in relation to the implementation of welfare-to-work policies in Australia. This unfolding policy agenda provides an opportunity, to examine how professional discretion is exercised in policy implementation within a statutory authority responsible for administering income support payments to hundreds of thousands of Australians. These issues are salient given the recent introduction of legislation to extend the welfare-to-work agenda to both sole parents and people with disabilities from July 2006. In implementing welfare reform Centrelink staff will have responsibility for interpreting guidelines, making assessments and administering activity and administrative breaches and suspensions. The implementation of this policy agenda raises important ethical issues about how front-line workers perceive their role and exercise discretion in implementing controversial policy.

A focus on professional discretion has its origins in the seminal work by Lipsky (1980) who proposed that street level work is an important, but often unrecognised locus of policy making. The literature on street-level bureaucracy includes arguments for and against the exercise of street level discretion and identifies the constraints within which discretion operates. Brodkin (2000: 7) argues that '...discretion is neither good nor bad but the "wild card" of implementation, likely to produce different results in different organizational contexts'. Brodkin's point about differing results is important for our focus as we examine the changing organisational structure and policy goals in the transition from welfare to workfare over the past decade. The first part of the paper briefly traces the policy and organisational context of welfare-to-work policies. The second part of the paper reports on previous research and preliminary findings from a current research project (1) examining professional discretion and the perceptions of Centrelink social workers in implementing welfare reform in Australia.

The Policy Context: From Welfare to Workfare

Australia's income support system has always had elements of 'mutual obligation'. However, the past fifteen to twenty years has seen a significant shift in Australia's income support provisions and employment services. The most significant shifts began with the 1989 Social Security" Review when the 'work' test was replaced by the 'activity' test (Ziguras et al., 2003). The change in the design and rationale of the income support system meant that social security was no longer a system of maintaining the incomes of those out of work; it became a system principally designed to encourage people into work.

As part of that program, employment services were reformed through the Keating Government's Working Nation policy package. Working Nation, introduced under the banner of 'reciprocal obligation', led to the Commonwealth Employment Service (CES) contracting out specialized training and labour market programs to non-governmental providers (Eardley et al., 2001). In 1996, the Liberal-National Coalition was elected into Federal Government. The change of government led to further reforms in the area of social security and employment services as reciprocal obligation was replaced by 'mutual obligation'--a more demanding form of 'obligation' on the part of the unemployed. The political justification for mutual obligation policy in Australia draws on 'New Paternalism' as articulated by Lawrence Mead (1997: 32):

 To live effectively, people need personal restraint to achieve
 their own long-run goals. In this sense, obligation is the
 precondition of freedom. Those who would be free must first be
 bound. And if people have not been effectively bound by functioning
 families and neighbourhoods in their formative years, government
 must attempt to provide the limits later.

Sawer (2000: 4) argues that mutual obligation policy merges neo-liberalism and new-paternalism as it is based on both contractual and paternalist rationales. The contractual rationale rests on the idea that the government provides financial support to people looking for work and in return, these people are morally required to put something back into their community. The paternalist rationale asserts that the unemployed benefit from participating in mutual obligation programs as it enhances their job prospects.

Since 1999 the Australian Government has sought to extend the principle and practice of mutual obligation to more groups within the community. From July 1, 2006 people in receipt of sole parent payments and those people with a disability deemed capable of work will be drawn into the government's narrowly defined 'participation agenda'. Workfare in Australia represents a fundamental re-fashioning of the regime choices Australia made in the early part of the 20th Century, and coupled with the dismantling of centralised wage fixing, the 'wage earners' welfare state' (Castles, 1985) has been rendered an historical artefact. We suggest that the transition to workfare in Australia represents significant institutional transformation (Phillips et al., 2004)--from the optimism of the post-World War 2 Keynesian welfare regime to a new regime characterised by New Paternalism and coercion (Carney, 2006).

In summary, the combination of reversing commitment to redistribution via labour market deregulation, the ongoing commitment to containing inflation at the expense of employment, the de-regulation of the labour market, the fiscal parsimony, and the linking of income security with active employment policy have all lead to the emergence of the workfare regime in Australia. Some immediate effects of the shift in rationality" at the level of policy implementation can already be discerned and it is to this level of policy reform that we now turn. We do so by following the impact on one group of specialist front-line workers--the Centrelink social workers. In the next section we examine the conditions under which policy conflicts have taken place in the implementation of the first round of 'mutual obligation' policies and the anticipated introduction of the latest round of welfare-to work policies.

From Social Security to Centrelink

In 1997, the then Department of Social Security (DSS) was transformed into Centrelink, a radical transformation in terms of Australian public administration (Rowlands, 1999). A pre-eminent example of New Public Management--inspired reform in Australia, Centrelink is the primary service delivery agency of the Commonwealth Government providing services on a contractual basis on behalf of an array of government 'client' departments and agencies. Because of its role in determining eligibility for various forms of social assistance (and the conditions that attach to them) Centrelink is a one-stop-shop monopoly (Keating, 2001). While Centrelink provides services to Australians, it is vital to remember that its clients are government entities; a design feature which in one stroke dissolved any vestige of income security as social rights of citizenship.

A driving rationale for the Government was a desire to improve the effectiveness as well as the efficiency of service delivery.. Adopting the language and style of the private sector, Centrelink management incorporated the latest devices for improving performance, for example, benchmarking and the 'balanced scorecard'. Consequently income support beneficiaries became 'customers' (though legally they are not) served by 'customer service officers' (once known as 'counter staff'). Offices were re-designed and an extensive array of 'customer' satisfaction" procedures were introduced, such as the 'value creation workshops' (ANAO, 2005), citizens' charters and so forth. The real efficiencies made by Centrelink have been through the extensive deployment of information and communications technologies (which, as we will see, has had a significant impact on the social work services). Given the size and complexity of the enterprise, there was little choice.

There are at least 600 social workers in social work-designated positions and a significant number in non-designated management positions (Fitzgibbon and Hargraves, 2001). In earlier versions of the agency the social work role was documented, for example, in a paper to the very first Australian Conference on Social Work (1947). Lyra Taylor, the first social worker appointed, said that social workers were employed to provide a:

'skilled casework service to the department's beneficiaries, to make the department's administration as humane as possible, and to form a useful instrument for social progress by assembling evidence on social questions' (Fitzgibbon and Hargrave, 2001, p. 123).

Further (and importantly) a community liaison and community service development role was articulated and extended in the 1960s and 1970s. Following a series of reviews, designated social workers were more closely articulated with the core business of the agency--income maintenance- but not in the sense of involvement in claimant eligibility determination. In 1996, the social work role was described as:

 To promote the wellbeing of Departmental clients by working towards
 the social justice objective of preventing and relieving hardship
 and suffering ... and to promote and facilitate the access of clients
 to Departmental income support programs and community, resources
 (Department of Social Security, 1996: 1).

The transformation from the Department of Social Security to Centrelink creates new sets of demands for all participants and social workers are no exception. In 1998, the Centrelink social workers were obliged to submit a 'service offer' to the Centrelink 'guiding coalition' about what they could contribute to the new organisation. Utilising the newly dominant reformist language, the Centrelink social workers did two things: first, they re-badged a large part of their traditional role within the rationality of New Public Management-inspired reform. Second, they identified at least two major 'spaces' for specialist social work practice ('fly away' work--responding to natural and man-made disasters, and 'virtual' social work- call centre work). In doing so, they quickly established competence and high degrees of acceptance and regard in both fields.

Eight years on from the submission of this service offer, what has been the impact of the new regime on the Centrelink social workers and their professional practices? Here we draw on two separate data sets, the first gathered in 2000 as part of a pilot project conducted in Victoria examining welfare reform at the front line, and the second generated as part of a national and ongoing project on social work and welfare reform in Centrelink. In regards to the latter, in 2005 an electronic questionnaire was circulated to the Centrelink social workers asking them about their daily practices. In all, eighty two responses from social workers located all over the country were received, predominantly in Customer Service Centres. By utilising both data sets we are able to focus on responses to the first wave of welfare reform from 1998 to the present and on social work responses to the anticipated introduction of the new welfare-to-work measures. The aim of looking at these two points in time is that it allows us to consider the extent to which the capacity to exercise professional discretion--the room to move--has been eroded.

The extent to which the logic and discourse of New-Paternalism and New Public Management have changed professional roles, social relations and organisational norms within Centrelink are important questions. While discourse is a significant site of struggle and change in organisations, we also need to consider the extent to which official discourse in written texts and talk changes the life-world discourse and actions of people involved. Analyses of organisational governance need to identify how much of contestation within organisations are (1) merely ripples on the surface of a settled modality of governance, (2) shifting parameters of established discourses and practice relations, and (3) unsettling the whole culture of governance relations (Healey et al, 2003: 67).

In the context of social policy practice we also need a satisfactory definition of policy practice that is sensitive to the relationship between discursive change and material social relations. The term 'practice' can be used to develop a unified way of knowing and doing; expressing the insight that knowledge cannot be separated from action. It would be wrong however to see practice as simply a synonym for action, given that policy practice can be as much about defending inaction and inertia, or '... insulating governance from the irrationalities of a mass public' (Torgerson, 2003:115). We need to find a way to connect the reflexive capacity of human agents to negotiate the discursive spaces that invite policy actors within organisations to '... take up the neo-liberally induced surveillance that holds us neatly packaged within economic and utilitarian discourses' (Davies, 2005: 7). We need to be attuned to the implications of a neo-liberal logic for both the unemployed-subjects and the professional-subjects of the welfare state. These implications extend to the unintended possibilities that emerge in day-to-day interactions when the invitation to be a 'responsible', 'independent', 'flexible' and 'self-reliant' self is refused or resisted.

In many organisations individuals regularly experience tension between their professional identity and the official organisational discourse. In the case of administering activity breaches in Centrelink, for example, social workers are likely to experience some conflict in their intention to support people's material and social wellbeing, while at the same time enforcing an organisational requirement to reduce their income support payments.

Resolving this tension may produce different forms of resistance. It is through a close analysis of these conflicts, both within and across organisational boundaries, that we start to understand what it means to individually and collectively challenge competition in social services, the transformation of clients into customers and subjecting expert professionals to market discipline and managerial discourse (Fraser, 2003: 168).

Tales from the front-line

When first introduced the breach penalty system attracted considerable public criticism from welfare agencies, lawyers and church leaders (ACOSS, 2001; Lackner, 2001; Hanover Welfare Services, 2000; National Welfare Rights Network, 1996; Salvation Army, 2001). A coalition of concerned agencies formed a taskforce and conducted what they called an Independent Inquiry into Breaching. Over a two year period, the Inquiry led by high profile lawyers and welfare advocates conducted its own investigations, interviewing clients and community agencies dealing with the effects of the policy. The Inquiry concluded by releasing a public report. The report concluded that:

 ... the current penalty regime is excessively harsh and unfair,
 and it unduly and counter-productively diminishes many jobseekers'
 prospects of finding employment (Pearce et al., 2002 p79).

The mobilisation of the welfare rights movement in the public sphere coincided with a concerted effort on the part of front-line social workers within Centrelink to have the harsh policies wound back. These organisational policy dissidents went about their campaign in a coordinated fashion, using available public service channels to feed their concerns up the management line, while drawing on the growing community concerns to support their case and open up a crack in the logic of the breaching regime. In this sense, the conditions for resistance and policy change resulted from an iterative process between policy 'insiders' and 'outsiders' building up a counter-discourse that briefly destabilised the government's 'getting tough' on welfare recipients discourse. In the words of a social worker who was involved in implementing the first wave of welfare reform in the late 1990s:

 Externally, the emergency relief networks, Welfare Rights
 organizations, peak bodies representing community services were
 voicing their concerns to the media and to members of parliament
 about the severe penalties imposed on those being breached. Two
 years later, the Minister officially announced changes to the
 breach policies to mitigate the harshness of breach actions and
 penalties. For me this example highlights how street-level
 bureaucrats will struggle to make effective policy changes without
 establishing connections with other internal networks (managers,
 unions) and external issue networks.

This excerpt draws attention to the conditions of resistance, which in this case were supported by policy, activism inside and outside Centrelink. The insiders--the dissident social workers within Centrelink--realised an opportunity for change when the 'community concerns' started to be brought to the public attention, as the following quote from another Centrelink social worker illustrates:

 As social workers, we were at the 'front-line' of the breaching
 policy's devastating impact on many vulnerable customers already
 with severe barriers to their participation in labour market
 programs. We were able to exercise discretion by waiving breach
 actions to several cases. Internally, the social workers led by its'
 National Manager in Canberra were one of the first organised groups
 to respond by providing evidence (case studies, stats etc) and by
 developing operational policies (e.g. no 3rd breaches without social
 workers approval). Many Centrelink staff who had to impose the
 breaches were also feeling the stress and were voicing their
 concerns to their union representatives. Senior level managers in
 Centrelink communicated the internal networks' concerns to the
 policy departments.

The post-script to this instance of discretion is that the institutional conditions that supported this action have changed over the last few years. The Department of Employment Workplace Relations has taken over much of the responsibility for the government's participation agenda (welfare reform) from the Department of Family and Community Services, particularly in developing the guidelines for the introduction of welfare-to-work measures on July 1, 2006. We now turn to interpretations of this intensification of welfare-to-work and its impact on social security administration at the coalface.

All respondents to the 2005 electronic survey reported an intensification of work in Centrelink over the past few years, not only for the social workers but also for the Customer Service Officers, with associated rises in anxieties and work-related stress. Within that, several other trends are clear. First, the people social workers are seeing are different from those they tended to see before Centrelink's inception. Mostly as a function of the participation agenda of the Government, and partially as a function of their role in assessing adverse participation reports (breaches), the service users they now see are increasingly categorised as having multiple problems (or in the language of the organisation, complex 'barriers to participation').

 We work more now with people who are disadvantaged by social
 isolation, estranged from family support, educational disadvantage,
 mild intellectual/behavioural disability, mental health issues,
 homelessness, drug and alcohol issues and severe financial
 hardship.

In the past, the DSS system rarely brought this cohort of income security recipients to formal Departmental or professional notice. Under the new conditions, the 'participation' requirements for accessing income support involve an increased degree of formal engagement and scrutiny. At the same time, the capacity for the social workers to respond in ways that they deem professionally appropriate (i.e. counselling and/or referral to specialist agencies) are increasingly constrained. In the first instance, they are constrained by the nature of their work processes (limited time and the requirement to maximise throughput).

 The new participation agenda crowds out time to do other important
 work such as crisis work or counselling. I have had to change how I
 practice ... from being a helping professional to holding a carrot
 and a stick.

In the second, the social workers feel constrained by what they perceive are increasing resource constraints in the community sector and state-based specialist agencies to which they would normally refer clients with complex problems for longer term intervention and/or treatment.

All survey respondents commented upon the centrality of 'participation' as the core rationality for guiding all Centrelink interactions with clients. The social workers, who at one time characterised themselves as generalist social workers (for example, engaging in assessment, brief and/or crisis interventions, referral and follow up, plus some community service development work with community sector agencies), now claim that they are required to project the need to 'participate' (in the workforce or in employment services) in all of their encounters with service users. In practice, this means that whereas once the social work service ran as a generalist social work service somewhat disengaged from the core administrative business of providing income support, under the current circumstances, promoting 'participation' has become core business, binding the social workers into the 'work first' agenda.

 It has had a dramatic impact on my everyday practice. No longer is
 this issue at the periphery of my practice but at the forefront.
 Also, Area Managers and others have made it abundantly clear that
 the focus of nearly all interventions must be participation.

One respondent noted that "participation is as much an ideology as a practice/ procedure", and in doing so, illustrated a degree of awareness and concern about the policy regime. Overall however, with very few exceptions, the social workers referred to the organisation's service users as 'customers' whereas once the dominant professional label would have been 'client'. This unreflexive use of the language of New Public Management sits somewhat at odds with the overall tenor or responses made by the social workers to the new policy regime. There was stronger resistance to the discourse of New Paternalism among social workers, while a greater degree of appropriation of New Public Management. By and large, the social work responses indicate dissatisfaction with the shift to workfare, as they are very concerned about the impact on the citizenship rights of people who engage with Centrelink (and it must be noted, on themselves).

 I can see a move in culture and a deliberate change in policy
 towards a more 'individualistic' view of citizenship. There is
 very little room for people to allow life to get in the way of them
 'participating'.

 The new participation rules are going to put families into
 frightening circumstances. Like a mother with a child with Cystic
 Fibrosis. Parents will have to decide "do I look after my child who
 may die if I don't do the care, or do I do what Centrelink wants
 and look for work so we don't get cut off payments and get food on
 the table?" Great choices for a prosperous society ... Coping with
 the injustice is a diff, cult emotional task for social workers
 these days.

Another theme constantly reiterated in relation to the effects of the 'participation' agenda is the subsequent narrowing or curtailment of professional autonomy. This takes multiple forms--increased accountability procedures, increased prescription in responses, more narrowly-defined expectations of what constitutes 'good outcomes', more detailed justifications for professional decisions. In regards to the latter, the social workers argue that their decision making in regards to adverse participation reports requires considerable justification, whereas when the policy of mutual obligation was first introduced it would have been accepted with less administrative scrutiny. The same applies to their use of extended exemption periods from participation requirements:

 My once liberal use of extended exemption periods in cases where
 people are experiencing social problems has, more recently, been
 curtailed. There is now case by case monitoring and surveillance of
 exemptions which are now required to adhere to strict and quite
 unreasonable time frames.

 The scrutiny of our reasons for "allowing" numbers of people in
 "non-active space", the expectation that there will be lots of
 intervention with people during this "time out" and that it will
 magically eliminate the social structural issues, so that within
 two weeks they will have secure affordable housing, or have
 safely escaped and resolved long term domestic violence so that
 they transform into marketable labour. These factors have so far
 been most influential in changing my practice--to the extent that I
 spend a lot more time explaining and justifying in "participation
 agenda' terms what I previously recommended in social work terms.

In summary, the traditional professional orientation to autonomy in practice has been systematically curtailed. The following quotation illustrates a widespread perception amongst survey respondents:

 I find that working in Centrelink, particularly with the
 participation focus, diminishes any potential for me to practice
 with initiative, limits my creativity in some ways and does not
 provide encouragement to enhance community contacts and
 development. Centrelink policies and practices do not focus on
 positive empowering processes. Instead they are directive and
 increasingly narrowly focused. Its impact is very damaging on the
 spirit of the customer and the community.

Some respondents, however, see the government's participation agenda in a more positive light. In general, respondents in this cluster (approximately 10% of this sample) argue that enhancing social participation has always been a primary focus for social workers. The new policy regime, they suggest, has brought some alignment between this traditional professional goal and the functioning of the policy regime within Centrelink itself.

 The 'participation agenda' has meant that people do not 'sit on a
 shelf and ignored. It means that we are beginning to follow up and
 to help long term until the person has moved forward--this, I think,
 is a wonderful change.

What this final quotation illustrates most usefully is that no one group responds to the intensification of welfare-to-work in unambiguous terms. It is as yet fairly early days in the broader project of welfare-to-work reform in Australia. The research referred to here tells part of the story and at a particular stage of its policy development. While the issues presented are clearly of interest to social workers, they are also important more broadly in that they reflect on how one group of people charged with policy implementation experience and anticipate further policy change. The disjuncture between a professional ethic of care, social citizenship rights and the narrow focus on paid work and training as evidence of 'participation' provides the conditions for policy conflicts within the organisation. A focus on these tensions reveals how organisational choices and decisions affect street-level practices and how street-level practices affect policy outcomes. That said, capacity for street-level practice to shape policy decisions is undoubtedly becoming more difficult. Under the new welfare-to-work regime individual social workers will be more constrained and have less individual responsibility for administering breach activities.

With the welfare-to-work changes to take effect from July 2006 Centrelink decisions about breaches and suspensions will be referred to an organisational unit separate from the Customer Service Centres with the managerial title of the 'Participation Solutions Team', which will be formally guided by the principles of efficiency and consistency. This development illustrates how the professional discretion of Centrelink social workers has been identified as a 'problem' by the 'purchasers' of Centrelink services (particularly by DEWR). We also witness how the discourse of New Public Management has become much more than a ripple on the surface of organisational governance; as in this instance it is fundamentally changing how the central tasks of this organisation are thought about and carried out. The new organisational configurations centralise decision making power on the issue of compliance and are likely to intensify the ambivalence that front-line social workers face in the new welfare-to-work regime. The implementation of the welfare-to-work agenda exemplifies the reformist dilemma of simultaneously being a bureaucratic agent of the state, a professional agent of the client and an ethical agent struggling to determine what is in the public interest.

Conclusion

By focusing on street-level social workers in Centrelink we can appreciate how policy is practised, including the factors that influence discretion, rather than simply seeing policy as delivered in accordance with formal policy objectives. While the empirical analysis we have presented here is preliminary, we think it makes a case for more research that examines how different forms of professional knowledge are championed and destabilised and how organisational configurations either close off or create possibilities for discretion. This evaluation approach places organisations at the centre of policy analysis and pays attention to both discourse and practice as sites of contestation at the local level. Traditional implementation research describes the extent to which managers and front-line staff carry out a policy with fidelity to its original intent and factors which impeded or facilitated that outcome. An organizational approach, by contrast, uses staff and target conceptions of their roles, their interests, and their cross-pressures and constraints as a starting point, examining how policy fares in the process (Brodkin, 2000).

Regarding the transition from welfare to workfare in the late 1990s we have presented evidence of 'critical professionals' inside Centrelink working alongside external social movements in fragile and fluid coalitions to effect policy change in the interest of protecting welfare rights and entitlements. In relation to the more recent ex-tension of the participation agenda and a narrowing of the social work role within Centrelink it appears that it will be more difficult to imagine and practice these creative possibilities for policy change. In other words, under the new regime there will be less room to move horizontally in terms of softening the harsh effects compliance measures in appropriate cases, and there will be less room to move vertically in terms of feeding information back up the formal decision-making chain. The intensification of welfare-to-work measures diminishes professional discretion for individual Centrelink social workers by designing away their room to move.

References

Australian National Audit Office (ANAO), 2005, Centrelink's Value Creation Program, Canberra, Australian National Audit Office.

Australian Council of Social Service (ACOSS), (2001), Breaching the safety net: The harsh impact of social security penalties, August, Sydney, ACOSS.

Brodkin, E. (2000) Investigating Policy's Practical Meaning: Street Level Research on Welfare Reform (working paper) [available online] http://www.jcpr.org/wpfiles/brodkin3.PDF?CFID=1413972&CFTOKEN=33782549 [accessed 26/04/06].

Carney, T. (2006). 'Welfare to Work; or work-discipline re-visited', Australian Journal of Social Issues, 41: 27-48.

Castles, F. (1985) The working class and welfare: reflections on the political development of the welfare state in Australia and New Zealand, 1890-1980, Sydney, Allen and Unwin.

Davies, B. (2005) The impossibility of intellectual work in neoliberal regimes. Discourse: studies in the cultural politics of education (26) 1: 1-14.

Department of Social Security (1996) Social Work Handbook, Canberra, AGPS.

Eardley, T, Abello, D. and Macdonald, H. (2001) Is the Job Network benefiting disadvantaged job seekers? Preliminary evidence from a study of non-profit job seekers, Sydney: SPRC.

Fitzgibbon, D. and Hargraves, D. (2001), 'Social Work in the Field of Income maintenance', in Social Work: Fields of Practice, (Eds.) M. Alston and J. McKinnon, Sydney, Oxford University Press: 122-135.

Fraser, N. (2003) 'From Discipline to Flexibilization? Rereading Foucault in the Shadow of Globalization, Constellations, Vol. 10, No. 2 pp. 160-171.

Hanover Welfare Services, (2000), Homelessness, income security and Centrelink breaching, Stats & Facts: A regular research report on Homelessness, December, Hanover, Melbourne.

Healey, P. and Magalhaes, C. Madanipour, A. and Pendlebury, J. (2003) 'Place, identity and local politics: analysing initiatives in deliberative governance', in M. Haajer, M. and Wagenaar, H. (eds) Deliberative Policy Analysis: Understanding Governance in the Network Society, Cambridge, Cambridge University Press.

Keating, M. (2001), 'Reshaping Service Delivery', in Are You Being Served: States, Citizens and Governance, (Eds.) G. Davis and P. Weller, Sydney, Allen and Unwin: 98-125.

Lackner, S. (2001), The unheard injustice: young people and Centrelink breaches, paper to the National Social Policy Conference, University of New South Wales.

Lipsky, M (1980) Street-level bureaucracy: dilemmas of the individual in public services, New York, Russell Sage Foundation.

Mead, L. (1997) (ed), The New Paternalism, Washington, Brookings Institute.

National Welfare Rights Network (1996), Penalties and poverty--I beg your pardon! An analysis of the severity, administration and impact of the penalties regime in the Australian social security system, August, Sydney, Welfare Rights.

Pearce, D., Disney, J, & Ridout, H. (2002), Making it work: The report of the Independent Review of Breaches and Penalties in the Social Security System, Sydney, Fineline.

Phillips. N. Lawrence, T.B. and Hardy, C. (2004), 'Discourse and Institutions', Academy of Management Review, 29, 4: 635-652.

Rose, N. (1993) Inventing Our Selves: Psychology, Power and Personhood, Cambridge, Cambridge University Press.

Rowlands, D. (1999), 'Institutional Change in the Australian Public Service--the case of DSS and Centrelink', Australian Social Policy, 1: 183-201.

Salvation Army, Australian Southern Territory, (2001), Stepping into the breach: A report on Centrelink breaching and emergency relief, August, Melbourne, Salvation Army.

Sawer, H. (2000), One fundamental value: Participants views on work for the dole, Unpublished working paper, Melbourne, RMIT.

Torgerson, D. (2003) Democracy through policy discourse, in M. Haajer, M. and Wagenaar, H. (eds) Deliberative Policy Analysis: Understanding Governance in the Network Society, Cambridge, Cambridge University Press.

Ziguras, S. Dufty, G. and Considine, M. (2003) Much obliged: Disadvantaged Job Seeker's experience of the mutual obligation regime, Melbourne: Brotherhood of St Lawrence, St Vincent de Paul Society and the Centre for Public Policy.

(Footnotes)

(1) This work is being undertaken by Catherine McDonald at The University of Queensland and Lesley Chenoweth at Griffith University. We thank Lesley for giving us permission to use the preliminary data.