Opinion article

Governments must be willing to show their spending works

It is crucial that government spending at all levels delivers results and represents value for money. CEDA analysis has found that over the past decade, 95 per cent of Federal Government programs with a total cost of more than $200 billion had not been appropriately evaluated. State governments fared similarly or even worse, writes CEDA Senior Economist Cassandra Winzar and Graduate Economist Sebastian Tofts-Len.

The Albanese Government has warned of “booby traps” left by the previous government as it prepares the public for tough decisions in next month’s Federal Budget.

With growing demand for key services and a shrinking tax base amid rising inflation, deficits continue to loom.

Although WA is in a much better Budget position, it cannot continue to rely on mining to generate large revenues to splash on programs.

In this environment, it is crucial that government spending at all levels delivers results and represents value for money.

But there is a way to ensure more spending hits the mark and to help avoid these booby traps — and we know it works.

State and Federal governments must up the game on evaluation.

CEDA analysis has found that over the past decade, 95 per cent of Federal Government programs with a total cost of more than $200 billion had not been appropriately evaluated. State governments fared similarly or even worse.

This should be concerning to all taxpayers. Hundreds of billions of dollars are being spent, with little understanding of whether it is effective or efficient.

We analysed 103 Federal, State and Territory program evaluations over the past decade in a range of policy areas, using auditor-general performance reports.

Of the 20 Federal programs analysed, a quarter (five of 20) had no evaluation framework and 70 per cent (14 of 20) had either an incomplete, inconsistent or poor evaluation framework.

The only major policy that we found had an effective evaluation framework was the administration of JobKeeper. This is likely due to it being relatively simple in its objectives and having good data sources. 

WA did not fare much better. CEDA analysis of 18 WA auditor-general performance audits in the past decade found 14 (78 per cent) programs had either no evaluation framework or had an evaluation framework with major issues. None had an effective framework.

WA has previously made some progress in this area.

In 2015, it became the first jurisdiction in Australia to establish an evidence-based policing unit to evaluate how to reduce crime through experimental trials. But the unit was axed in 2020.

WA’s Target 120 program, which aims to direct at-risk youth away from the justice system, has a strong focus on progressive evaluation and data sharing to inform the ongoing development of the program. It is showing some promising signs.

The Economic Regulation Authority, WA’s independent economic regulator, has recommended that election commitments should be subject to rigorous evaluation before they are included in the State Budget, and the outcomes of all major project evaluations should be published.

This recommendation should be implemented.

The current system encourages policymakers to take rapid responses to social problems, rather than insisting on regular evaluation of existing programs.

This is compounded by challenges in undertaking evaluations and a lack of resourcing. Plus, good evaluation is hard.

The Albanese Government has proposed an Office of the Evaluator-General to kick start improvements in this area.

This could be announced in the upcoming Budget.

Detail has been scant, but it would be an important first step.

CEDA believes an OEG will need a clear remit to be successful.

We propose that its chief role would be to champion evaluation and develop capability and capacity throughout the public service.

Evaluations would continue to be primarily conducted at the departmental level, however, there would now be a body stewarding and championing evaluation.

An evaluator-general is not the solution to all problems with assessing government programs. But it is the kind of circuit-breaker needed to drive change.

Leveraging the evaluator-general model, the Federal Government must also address incentives, improve evaluation practices, invest in data availability and access and improve governance practices.

This cannot be left to the whims of departments or ministers. The Government should legislate the review of existing major Commonwealth-funded programs at least every five years.

While evaluation is important across all government policy areas, it is in the community services designed to tackle entrenched disadvantage where we consistently see the failings of poor policy and limited or no improvements in service delivery. The Government must begin its commitment to regular, robust evaluation here.

This article was originally published in The West Australian on 28 April 2023.

About the authors

Sebastian Tofts-Len

See all articles
Sebastian Tofts-Len commenced his professional career with CEDA in 2022 as its new Graduate Economist, based in WA. During his undergraduate studies, Sebastian undertook internships with the Chamber of Commerce and Industry WA, as well as the Mannkal Economic Education Foundation. He also worked as a research assistant at Curtin University, drafting literature reviews on the economics of climate change. Sebastian holds a Bachelor of Commerce (Economics) with Distinction from Curtin University.

Cassandra Winzar

See all articles
Cassandra Winzar is Senior Economist (WA) at the Committee for Economic Development of Australia (CEDA). Prior to joining CEDA she was Principal Economist at the WA Department of Communities (Housing Authority) where she focused on WA economic conditions and housing related research, including running the state government’s Housing Industry Forecasting Group. Cassandra has also held roles as the WA based Economist for the Reserve Bank of Australia, and in Transfer Pricing at EY. Cassandra has a Bachelor of Economics (Honours) and Bachelor of Arts (Asian Studies) from the University of Western Australia.