An overarching strength of the CDEP scheme, as currently constituted, is its flexibility. It is at once an employment, training, community development and enterprise assistance program. At times it is just an income support mechanism, little different from welfare. The popularity of the scheme is evident in its spread across 290 communities Australia-wide, in very diverse circumstances. This flexibility is also a weakness, because while there is statistical evidence of success in employment and income generation, it is unclear how successful the scheme is in community development and training. This is partly because the latter are more difficult to measure, and partly because in many cases appropriate statistics do not exist. Because of the diversity of its program objectives, CDEP scheme performance is hard to measure and its success, where evident, is difficult to demonstrate.

In many situations CDEP organisations try to do too much, especially those that are successful. This is often because other community-based specialist agencies are non-existent, and as a result CDEP support provided by ATSIC sometimes substitutes for the programs that should be provided by other agencies. If CDEP organisations are to do the jobs of other agencies, be they DEWRSB, DETYA, Centrelink, or other Commonwealth or State agencies, it should be on a full cost-recovery basis (see comments in this volume by Bartlett, Ch. 20; Kean, Ch. 21; Nalliah, Ch. 28). Similarly, if CDEP organisations contract to undertake municipal services or operate as outstation resource agencies, this should be done on a commercial basis and be funded accordingly. There are existing best-practice examples of such approaches.

Ultimately, if CDEP organisations are to focus on employment and community economic development in the new social policy environment, it might be appropriate for the scheme's objectives to be amended and negotiated on a case-by-case basis. There are dangers inherent in pursuing too many diverse objectives.

As noted above, McClure (2000) provides few recommendations or solutions for the development problems of remote regions, beyond noting options for establishing community development institutions. Rather than establishing new and additional organisations, it would be far better to use an existing institutional model, the CDEP organisation, to deliver development opportunities at the community level. Rather than inventing new institutions from nothing, it would be more cost-effective to resource CDEP organisations to build their capacity to deliver development. Interestingly, Pearson (2000b) does not nominate CDEP organisations as the governance structures to facilitate development, recommending instead larger regional agencies. More recently he has favoured direct relations between agencies and families and individuals (Pearson 2000a). Intermediary community-based organisations will undoubtedly be needed to facilitate development: the question is whether, as is argued here, CDEP organisations are the best placed to become development agencies.

The answer depends in part on resourcing. In some regions, ATSIC has provided organisations with triennial funding on preparation of development plans. This has allowed a degree of capital and operational funding commitment that has greatly assisted planning for development. A danger that some organisations face is that success might reduce discretionary capital funding; it is important that this does not occur and that funding decisions do not become politicised and biased against the successful.

An advantage of the CDEP as a development agency model is that CDEP organisations can be constituted in a way that separates trading activity from ATSIC-supported activity. This in turn means that trading profits—members' rather than public finance—can be invested in high-risk ventures.

To make the transition to their role as development agencies, CDEP organisations will require resourcing for board and member empowerment. In particular, there is a need for an additional injection of funds for training boards and for participatory planning. If CDEP organisations are to change into 'community economic development' organisations it is imperative that their growing complexity does not result in the alienation and disengagement of their members and over-reliance on non-Indigenous management. Such disengagement can result in organisational instability that is not conducive to sound business practice.

There are mismatches between the CDEP scheme and mutual obligation, and a logical tension between welfare and CDEP regimes, especially when they are found in the same community. For example, as a general rule, CDEP combines high activity testing with almost no income testing. Paradoxically, New Start Allowance (unemployment benefit) in remote situations has almost no, or highly modified, activity testing.^[23] Because its income test is so much more generous, CDEP provides positive signals to participants to supplement their incomes either in monetised or in non-monetised sectors, in a way that welfare does not.

It will be difficult to counteract the emerging fusion between CDEP and welfare institutions and the mixed signals that ensue. CDEP 'top up' provided by Centrelink to provide equity with the ‘work for the dole’ program should be provided to CDEP organisations to administer. The administratively-burdensome CDEPManager electronic database, and daily monitoring of CDEP participation (which far exceeds Centrelink monitoring of New Start Allowance clients), should be reconsidered: CDEP organisations should be allowed flexibility in differentiating types of participants and should be allowed use of wages surpluses for top up. Finally, to be effective and to encourage diversion from passive welfare, CDEP organisations should be provided with wages, capital, and operational support budgets that are open-ended, bearing in mind that most cost (76 per cent according to 2000–01 estimates) will be nominally offset by welfare. A discussion of the means to make these funding changes falls outside the scope of this paper. Delivering development in remote Australia will not be cheap, but the utilisation of a scheme that has considerable welfare offset is a relatively cheap option for government.^[24]