Concerns about implementation capacity have been reinforced by high profile implementation and delivery failures. In Canberra, the treatment of two mentally ill Australian citizens wrongly deported by the Immigration Department became a major scandal, precipitating wide-ranging reforms to the structure and culture of the agency (Palmer 2005). The botched repatriation of the body of Australian soldier, Jake Kovco, accidentally killed in Iraq, exposed coordination difficulties in the interface between the Department of Defence and private contractors, at the cost of great hurt to the bereaved family, and major embarrassment to the Defence Minister and the government. In Queensland, a litany of delivery problems has plagued the Beattie Labor government during its third term, most seriously the ‘crisis’ in the State’s public hospital system (Tiernan 2006). These cases have served as ominous reminders to politicians and public administrators alike of the serious political costs and consequences of implementation failure. As Peter Shergold noted recently:

Much of the focus of the Australian implementation units has been on ensuring delivery issues are planned and addressed in new policy areas. The problems and failures afflicting the two governments have occurred in established service delivery systems. In the Immigration and Queensland hospitals cases, governments responded by establishing independent inquiries, including in the Queensland case, a royal commission. As well as providing forums for investigating the factors that led to the events in question, the reports of these inquiries have generated useful blueprints for reform and change. They have highlighted a serious disconnect between policy and service delivery – the classic implementation deficit identified by Pressman and Wildavsky (1973) in their seminal study of implementation failure.

In Queensland, consultant Peter Forster (2005) who led the independent review of public hospitals, was particularly critical of central agency and head office officials for failing to appreciate and address systemic problems of under-funding, workload issues, and the difficulties of recruiting and retaining appropriately qualified staff to work in the state’s public hospital system. Forster (2005) describes a major ‘expectation gap’ between what politicians and the public expect can be delivered and what service systems are actually capable of. Managing public expectations is an invidious and likely insoluble implementation dilemma for politicians, as recent criticism of the Commonwealth and Queensland government responses to the Cyclone Larry disaster have again demonstrated. Despite a swift and focused emergency response to widespread damage wreaked by the cyclone, and the difficulties of establishing services in the absence of electricity and transport access, governments were criticised by some locals, frustrated by delays in gaining access to relief supplies and funds. Their complaints were amplified through the broadcast media – a young woman’s anger at perceived inaction of ‘bureaucrats’, broadcast to an attentive national audience.

The Immigration and Queensland hospitals cases also highlight the difficulties of ensuring that operational realities are reflected in policy advice and decision-making, though it is interesting to note that both have been identified primarily as public service failures – as failures of analysis and persuasion (see Briggs 2005; Shergold 2006b), rather than failures on the part of policy-makers to understand systemic problems and direct energy, attention and resources towards addressing them before they escalated.

In the wake of these controversies, the Australian and Queensland governments have initiated wholesale changes to affected departments. In Immigration and in Queensland Health, the entire senior leadership teams were replaced, and the organisations were radically restructured. There have been major funding injections, and agency-based implementation units are bolstering the commitment to ‘fix’ the identified problems. Strong central monitoring and reporting arrangements have been established, including requirements to provide regular reports to Parliament. ^[3]

But rather than building local capacity, the appointment of significant numbers of central agency staff to leadership positions in agencies like Immigration and Queensland Health raises questions about the value placed on content knowledge and service delivery expertise. Though perhaps understandable in an increasingly personalised governance context, there are tensions between building capacity for implementation and parachuting it in. It may further undermine confidence in agencies already regarded as having failed in their duties to government and the community. In both cases, delivery problems have persisted after some initial blood-letting; agencies and their responsible ministers remain in the media spotlight, as new leaders try to bed down hastily devised political ‘fixes’ and confront the very genuine complexities of large-scale system reform.

These developments reflect the inherent tensions between ‘the normative expectations of managerial control of policy implementation processes’ (Barrett 2004, p. 260) and the realities of implementation in a networked and highly politicised service delivery context. Tiernan (2006) notes the predominance of a ‘top-down’ view of implementation in the development of implementation units and their monitoring strategies, especially in areas that have caused political embarrassment. Barrett (2004) and Hudson (2006) observe a similarly top-down orientation to implementation in the British context, noting ‘there is a lack of recognition of the time and resources involved in achieving the organisational capacity to achieve effective change’ (Barrett 2004, p. 260).

Political pressure to quickly address problems and failures in sensitive areas of public policy may have perverse unintended consequences, potentially embedding new and different implementation challenges down the track. For example, Barrett (2004, p. 260) describes how ‘top down coercive pressure to meet prescribed targets’ has ‘led to the skewing of service priorities’ (in this case hospital waiting lists), and ‘even the manipulation of figures for the fear of the consequences of failure’. Shergold, however, is unapologetic about the need for the focus on implementation and delivery to be driven centrally. He argues that ‘better implementation must consciously be driven from the top down’, while simultaneously acknowledging that policy prepared without the experience of those who deliver it ‘is almost certainly policy that will be poorly designed and difficult to implement’ (Shergold 2006b, p. 3). Reconciling these competing imperatives would seem to be the key challenge facing governments in building capacity for policy implementation.

What role will implementation units play in this agenda, and how and in what ways will their establishment help to address the problem of implementation capacity in contemporary government? What are the prospects of addressing the capacity problem through the establishment of implementation units? The papers contributed to the special issue of the Journal of Comparative Policy Analysis (JCPA) on the emergence of implementation units demonstrate clearly that some progress has been made. Central units are playing leadership roles in promoting and raising awareness of implementation at the ‘front end’ – during policy development and design, though there are issues about their capacity to do so. As fairly small units, whose work is closely linked to the strictures of the Cabinet timetable, and driven by the priorities of the first minister, central implementation units have limited capacity to undertake the kind of outreach activities that would help to build capacity for implementation across government. Peter Shergold (2006b, p. 3) describes these functions as being about ‘learning by doing and then spreading the learning’.

The Immigration and Queensland hospitals controversies suggest that if the goal of the new focus on implementation is capacity building, a more constructive role for central units may be in helping promote better understandings of implementation issues and challenges, particularly among decision-makers. There is also an important role to be played in assisting agencies to more effectively communicate the complex realities of translating decisions into actionable programs to policy-makers. Greater engagement with the literature on policy implementation and the policy-action relationship (Barrett 2004) could be a useful first step, yet Lindquist (2006) reports there has been limited engagement with the implementation literature in the design and development of implementation units. In addition to their enthusiasm for project management techniques, those interested in building capacity for policy implementation would do well to revisit the scholarship and adapt some of its learnings to contemporary practice.