Does tree-shaking work? Evidence based policy and welfare conditionality

This article from Dr Simone Casey explores why Australia’s Mutual Obligation requirements are so demanding and whether this is based on evidence about what works. It asks why critical research evidence has not received more attention from Australia ‘s activation policy makers. She argues that lack of engagement with critical social research is a limitation which hampers social justice efforts and reflects disregard for social suffering, and says there is plenty of room for stronger engagement with participatory policy design approaches. Dr Casey is an Associate of the RMIT Future Social Services Institute.

Photo by  Trung Thanh  on  Unsplash

Photo by Trung Thanh on Unsplash

Activation of the unemployed has accompanied labour market program reforms since the 1980s. Since the influence of neoliberalism there have been shifts to market-based welfare interventions at both individual and institutional level. Labour market programs produce market-consistent behaviour in the unemployed intended to commodify them for the supply of contingent labour and to rationalise welfare expenditure. 

For unemployed citizens this shift to market relations has involved activation and welfare conditionality. If activity levels are not met “job seekers” are subject to financial sanctions premised on contestable behavioural economic theory about motivation and behavioural change. Mutual obligation (MO) is the term used to describe these relations - it is a variant of the personal responsibility discourse appropriated to capture the contractarian nature of the welfare contract in Australia. 

There have been recent significant changes to MO in Australia that have not received public scrutiny. Work for the Dole, which was the first introduced in 1997, is a somewhat unique feature of Australia’s mutual obligation regime which obliges job seekers to participate in unpaid work at a community run organisation. Jobactive services are required to enforce this through implementing a phase of service called the Annual Activity Requirement (AAR) for six months of every year of unemployment.  At the start of jobactive the AAR for adults aged 30-50 it was 15 hours per week.  

However, in September 2018 the requirement increased to 25 hoursper week. This is despite the OECD already having identifiedour activity levels as amongst the highest in the world. These high levels of activation combined with low unemployment payments have imposed onerous conditionality on vulnerable job seekers.  

These increases in MO hours are significant, yet appear to run counter to the accrued body of evidence about what works in recent welfare conditionality research. Demanding activationand low job availability leads to demoralisation and has been challenged as a human rights issue. The OECD warnedthat governments should aim to strike a balance between demanding activation measures and the need for universal social security and these concerns were repeated in a systematic review in the UK. Effects such as withdrawal from income support are of concern because they mean vulnerable people may not access services that could help them. 

Closer to home, Australian studieshave shown that financial sanctions are distressing for people already surviving on low incomes, making it harder for them to meet the costs of looking for work, increasing their dependency on kinship networks or increasing the likelihood of criminal activity. These findings are similar to those of the UK welfare conditionalityproject which identified undesirable consequences, such as ‘distancing people from support; causing hardship and even destitution; displacing rather than resolving issues such as street homelessness and anti-social behaviour; and negative impacts on third parties, particularly children.

Furthermore, evaluations of Work for the Dolehave shown it is not effective as a labour market intervention.  This is because of the negative interaction between coercive policy interventions and human motivation; and that ‘make work’ programs should be sensitive to labour market conditions and linked to the potential for job creation.

Photo by  Joanjo Pavon  on  Unsplash

Photo by Joanjo Pavon on Unsplash

Evidence-based policy?

So what is the evidence underpinning this policy? It seems counter-intuitive to impose conditions on job seekers that reduce their availability for genuine job search, and which run the risk of being so onerous they exit from social security. Perhaps the answer is that bureaucratic evaluation is little more than cost-benefit analysis as departments seek to justify funding for their programs.  This means the analysis focuses narrowly on the terms and mode of delivery rather than on the subjective effects of activation and welfare conditionality. Perhaps even more concerning is that this perpetuates an assumption that doing more of what they’ve already done is the only way to go.

Department evaluations can be classified as either administrative evaluations of the outcomes generated under different contract models or are of activation strategies such as the Active Participation Model (APM) and Welfare to Work. These evaluations use empirical measures such as the number of job outcomes, cost benefit and statistical measures of job seeker post-program outcomes and satisfaction. However , this type of evaluation has been subject to criticism because the results mask outcomes at individual levels as levels of disadvantage in the labour market have increased. The OECD has also suggestedthat program evaluation that looks beyond the 26 week outcomes of employment assistance programs would provide a better indication of their long term benefits.

Further, it has been argued that MO activities like work for the dole have a tree-shaking effect. Recent evidence provided by the Department of Jobs and Small Business (DJSB) to the Senate Estimatesfor May 2019 suggests that tree-shaking is an active strategy because they anticipate 20 and 30 per cent of people go off benefit when they enter the work for the dole phase.  This admission suggests tree-shaking is considered good policy in Canberra. In light of this, it is unsurprising that representatives from the department recently claimed the UK welfare conditionality project findings were based on flawed methodology.

The evidence base used in this policy ignores the concerns of citizens. Even though the Senate Inquiry into jobactive highlighted significant concerns about mutual obligation, activation and work for the dole, these matters were excluded from the terms of reference for the future employment services reforms. An official evaluation of changes to MO is currently underway without input from peak bodies or consumer advocates.  Yet one of the most striking matters to have emerged from the recent Senate Inquiries into jobactive and ParentsNext was the failure of DJSB to engage in genuine consultation on program design with citizen consumer peak bodies such as the Australian Unemployed Workers Union or the National Council for Single Mothers and their children. 

To pay more than lip service to consultation, far more radical evaluation strategies should be employed. Interpretive policy studies can be used to understand the subjective impact of policy on human subjects, driven by humanist and emancipatory goals rather than economic/cost benefit. These interpretive approaches apply critical insight into policy making and the frameworks that are used evaluate program effectiveness. 

Photo by  Joanjo Pavon  on  Unsplash

Photo by Joanjo Pavon on Unsplash

Genuine participation is urgently needed

The 2018 decision to increase MO hours appears to have been made in a vacuum with little critical engagement with studies that have warned of the consequences of demanding activation. It should by now be clear there is a problem with the extent to which real consultation has occurred with citizen advisors in policy design. It is important that policy makers rectify this by ensuring the concerns of representative bodies such as the AUWU are genuinely included in future reform processes.