Friday, January 22, 2010

Why Jenny Maklin's standards approach to child welfare is likely to fail

My heart sank as I read newspaper reports about the latest Federal Government moves on national child welfare standards, more precisely  on those living in care. To quote Minister Jenny Macklin:

We need national standards of care so children who cannot live with their families can grow up in a safe, secure environment. National standards will provide a benchmark for the care of these children no matter where in Australia they live.

Before going on, you can find the Minister's press release here, the standards consultation document here, the Australian Institute of Health and Welfare report here.

Why did my heart sink? To my mind, it is yet another policy move doomed to almost certain failure, another example of the weaknesses in Australian public administration that I started to address in a series of post last year. I have not pursued this series as actively as I had intended in part because of time, in part because I became depressed. Now I need to return to the issue.

My argument in the series was a simple one. Current Australian approaches to public administration do not and cannot work. To establish this I had to demonstrate a pattern of systemic failure, to establish the reasons for that failure, and then to suggest alternative approaches. In the first instance, I adopted a case study approach.

In considering the latest Federal moves, I think that we need to begin by recognising two things.

The first is that kids are placed in care because they are experiencing personal problems. These may be behavioural or, more likely, the family circumstances threaten the child in some way. The first and key judgment to be made as to the success or otherwise of the placement is whether or not the kid is in fact better off, more broadly whether the totality of kids in care are better off. The question of how much better off is a further question to be asked once the first is answered.

The second is that all child welfare systems throughout Australia have been experiencing problems. I have written especially on the NSW system. There the loads and expectations placed upon the system by things such as mandatory reporting brought it to the point of collapse. So there is a problem.

The idea that national standards of themselves might bring benefits can be a very slippery one, in part because there is a lot of confusion over the meaning of the word standard.

Of itself, a standard has nothing to do with good or bad. The focus is on fitness for purpose. A standard is designed to achieve uniformity, to ensure that the product or process does a specified set of things or meets a specified set of criteria. So in child welfare terms, a national standard for children in care means specifying a set of things that the child welfare process must deliver for those children in care.

A standard is fixed. It says nothing about the value of alternative standards, nor does it deal with cases that may exceed the standard in some way. It just is, freezing certain things at a point. Here standards often impose costs because they preclude alternatives. The usual justification for a standard is that the benefits from standardisation exceed those costs. This often is, but may not be true.

Compliance costs are one of the of the standard standards' costs. This is especially true for process standards, including the attempt to create management standards. Whole systems have to be set up to enforce the standard. In management standards, this includes process material along with compliance measurement. One of the reasons that the fashion for management standards peaked in the 1990s is that the compliance costs proved to be high, the actual benefits lower than expected. 

Against this background, the success or otherwise of a national child welfare standard rests on the assumptions that setting a specified set of things that the child welfare systems across the country must deliver to children in care will deliver a better set of results taking compliance costs into account than an alternative process. This is, in fact, quite uncertain.

Problems begin with the definition of the standard itself, the purpose of the current discussion paper.

Take, as an example, the inclusion in the discussion paper of the need to recognise Aboriginal cultural values. This is very much current public service think.

Those who read this blog will know that I write a lot on Aboriginal issues and policy, including the need to focus on Aboriginal history and the Aborigines as varied peoples. I make this point only because I fear that I am about to get into serious trouble.

Aboriginal children are many times more likely to end up in care than non-Aboriginal kids, reflecting social dislocation within the Aboriginal community. It is, I think, current policy in all jurisdictions that Aboriginal kids are better off placed with Aboriginal families for cultural reasons. This policy has already led to some tragedies and has bound the hands of child welfare workers, forcing them into what they know are second best solutions.

To my mind, the sole issue is whether or not kids at risk would be better off in care. The question of Aboriginal cultural issues is one thing to be considered in the context of this primary aim.    

Similar issues arise with so called CALD kids, culturally and linguistically diverse, another example of modern public service speak.

Problems become more intense when we look at the detail of the areas to be addressed by the proposed standards. They simply mix together too many things. They can also conflict. I am not saying that we should not have them - who could argue with regular medical checks? - just that they lack clarity and balance considering the primary aim.

Generalised standards areas then have to be translated into measurable standards. This is usually done via some form of statistical measure. Another set of problems arise.

To begin with, the standards are system wide standards. I have no problem with system standards. However, they generally do not recognise variance within the system, the extent to which the system actually fails individual children. Here a system may meet broad standards yet in fact be failing individual children worse than another, apparently better performing, system. In theory you can set up measures to accommodate this, but the process is reasonably complicated.

Then, too, the measurement of standards performance at national level is strongly influenced by the availability of statistical data. The relationship between the data and the actual on-ground position can be quite uncertain. This is complicated by a further variable, the nature of linkages and relationships between elements within the system.

Consider this as a simple example. One jurisdiction may in fact take a higher proportion of children into care than another jurisdiction. Its stats will show this and may well, depending on the way standards are measured and defined, count as a negative. However, that jurisdiction may also show better performance for children in care than other jurisdictions because the proportion of children with very serious problems is less.

Assuming that all these problems can be overcome, at least two further problems arise.

The first links to the question of compliance costs, including jurisdictional responses to problems as they arise.

The collection and presentation of data to meet national reporting requirements has its own costs. These costs link to the initial definition phase and then to the reporting and compliance stage and are not insignificant. Unfortunately, the data so collected is rarely useful from an operational viewpoint. It is simply too generalised.

The standards and data reporting requirements also twist agency responses to problems as they arise. They impose a need to respond to shortfalls as measured by the standards or indicators involved. These may but need not be linked to key delivery challenges as they stand at the time.

The final problem is simply one of resources. There is no point in setting either standards or targets if the resources are then not available to ensure delivery. Even if the target or standard is in fact achievable, and that is often not the case, it will fail if resources cannot be found. We have seen this happen many times; it is a major reason for policy failure.

While I think that the approach is systemically flawed, I would be more confident if Minister Macklin and her Department had a better record of service delivery. I dealt with one aspect of this in Jenny Macklin and problems in NT indigenous housing.

So what do I expect to happen in all this?

I really hope that I am wrong, but my feeling is that a lot of time and resources are going to be devoted to the standard setting exercise and then to reporting and responding without a single tangible gain where it matters most, improved care for Australian kids.       

No comments: