Assessing the 'radical experiment': the Job Network after four years

When the Job Network began operating in May 1998, replacing the former Commonwealth Employment Service and contracted-out case management agencies, it was variously described as a 'radical experiment' and a 'revolution in employment services'. Now four and a half years old, this quasi-market in employment services is to undergo another transformation to fit the new 'Active Participation Model', itself devised to reflect what has been learned about the Network's successes and failures over the past four years.

Until recently there has not been much detailed information available about the workings of the Job Network and it is still the case that independent researchers are unable to access much of the original data on which official evaluations are based. However, the Productivity Commission's final report on the Job Network has now been published, along with the Government's response. The Department of Employment and Workplace Relations' third evaluation report was also released at the same time, with interesting new data on net employment impacts. It is therefore a good time to take stock of what we now know about the radical experiment and to consider what is promised under the new model from mid-2003 onwards.

Job Network Outcomes

The latest reports provide a mixed picture of the Job Network's achievements. Overall they confirm earlier estimates that outcomes, while gradually improving, appear broadly similar to those of the previous Working Nation programs at their peak, taking into account differences in program participants, the external employment environment and methods of assessment. On the other hand, the Productivity Commission also confirms what earlier research, including our own, has suggested, which is that weaknesses in the funding structure have led to widespread 'parking' of harder-to-place job seekers, many of whom receive little help while in the intensive phase of assistance. There are also clearly still problems of access and commencement of assistance for some key groups, including young people and Indigenous job seekers. A number of submissions to the Productivity Commission's review highlighted what they saw as a highly fragmented employment system for young people and a lack of "joined-up solutions" to their particular difficulties.

As the Productivity Commission notes, gross outcome measures can be misleading and tell us little about the net impact the employment programs have on unemployment. Previous net impact measures have been based on departmental post-programme monitoring studies. These have suggested, for example, that the prospects of Intensive Assistance leaving income support were around 10 per cent greater than those of a matched comparison group. However, both the OECD in their review of the Australian labour market and the Productivity Commission have highlighted methodological problems in the way the Department's post-programming monitoring studies have been designed, which tend to exaggerate the impacts. This has been recognised in the Department’s recent evaluation of the net impact of Job Network services, which is based on new methods that take better account of three factors that are likely to be important.

These factors are the compliance effect – the effect on job seekers' behaviour from simply being referred to a program; the program effect – the actual benefit gained from the services provided on the program; and the attachment effect – the generally negative impact of reduced job search effort while involved in a program. On this basis, the Department now estimates that the average net impacts of assistance are much smaller than previously suggested – just over two per cent for Intensive Assistance and between seven and eight per cent for Job Search Training. Most of the impact (virtually all in the case of Intensive Assistance) seems to be achieved not through program assistance itself, but as a result of people moving off income support simply by being referred to the programs. The assumption in the latter case is that when referred, people either increase their own job search effort or are already working unofficially.

The results are perhaps not as bad as they might seem at first sight. The new estimation methods are tentative and probably conservative, while the averages conceal variations that suggest significantly higher net impacts for more disadvantaged job seekers. Overall the results are not out of line with international experience of the relatively small net employment gains to be had from labour-market programs and are also probably similar to those achieved under previous assistance arrangements.

Cost Effectiveness?

These smaller net employment gains bring into question some of the Government's bolder claims of greatly improved cost effectiveness under the Job Network. DEWR comparisons of the costs of both gross employment outcomes and net off-benefit impacts (based on their previous methods) from Job Network and Working Nation programs suggest that the former are substantially lower on average. The newer and more modest net impact estimates, however, make Intensive Assistance in particular look a lot more expensive per net employment gain. Similar caveats would also apply to the net impacts of Working Nation programs of course, but as the Productivity Commission states in noting that the aggregate cost of labour market programs fell by around half between 1996-97 and 1999-00 without much difference in unemployment levels: "This may well be the result of greater cost effectiveness of the programs, but it could also be the result of the imprecision with which the small impacts of labour market programs are measured."

There seems little doubt that competition within the framework of Job Network has produced some efficiency gains. It is difficult to pinpoint quite how such gains are achieved because it remains hard to find out much about what different agencies are doing on the ground – too much has been 'commercial in confidence'. But some examples include innovative relationship building between some agencies and employers and other bodies to develop targeted employment opportunities, and the flexible use of outcome funds to meet the range of costs for effective job search and preparation. There was undoubtedly less flexibility for such approaches under the CES.

On the other hand, competition among service providers has had other spin-offs that are not easily estimated and are not captured at all in the simple focus on measured employment outcomes. For example, a number of providers and individuals consulted by the Productivity Commission, as well as in our own study, pointed to some negative impacts of competition on relations within the not-for-profit sector and on communities where long-established local agencies ran into financial difficulties in the earlier rounds of the Network. We also have to bear in mind employment losses, as well as reduced pay levels and stress among some placement staff in the privatisation process. The apparent efficiency gains of getting broadly similar outcomes for a much reduced public expenditure dollar have to be set against some of these displaced costs, which are not well understood at present.

Sustainability of Employment Outcomes

One of the other key questions that needs to be asked in relation to recorded outcomes of employment programs is whether the jobs gained last much longer than the time required to produce a paid outcome. Sustainability of employment for disadvantaged job seekers has become an important goal of policy in countries including the UK and US, in order to avoid the problem of recycling through employment programs. The DEWR evaluation goes some way to addressing this issue through longitudinal analyses of outcomes from different Job Network programs. It suggests that many of the jobs achieved do last longer than might have been expected. It also points to some upward movement over time in terms of wages and hours of work on the part of Job Network clients – though this trend is less marked among those who began with relatively low-quality jobs.

Policy Dilemmas

The twin problems of ‘parking’ (a reluctance to put effort or expense into harder-to-place clients) and 'deadweight' (the high proportion of positive outcomes likely to have happened even without assistance) present serious policy dilemmas. On the one hand, the research outcomes suggest the need for a higher threshold of disadvantage for access to Intensive Assistance, in order to reduce the amount of deadweight in the system. On the other hand, there is the equally pressing question of how to help the already more than 60 per cent who receive such assistance without getting an outcome – as well as those coming round for a second time. The question of how to deal with these problems marks a point of distinction between the different responses of the Productivity Commission and the Government.

Overall, the Productivity Commission has, perhaps not surprisingly, opted in favour of greater competition and a purer market-based solution. It suggests deregulating the ways in which providers are able to assist their clients by giving them more freedom to refer on those with greater employment barriers to other programs (such as community work or Work for the Dole). There was strong opposition to this proposal from many of those responding to the Productivity Commission's draft report and the Government has not accepted it at present. However, one way of looking at restructuring the former Community Support Program into the expanded Personal Support Program under the Department of Family and Community Services is in terms of movement towards focusing Job Network services more on those with the capacity to benefit.