Menu Close

Getting more bang for public bucks: is the ‘efficiency dividend’ efficient?

In the 27 years since the Hawke government came up with a public service efficiency dividend, the evidence has mounted against it. National Archives of Australia, CC BY

Every now and again in public policy debates a consensus emerges on some particular point among policymakers, stakeholders and commentators. These moments are distressingly rare. It is even more distressing when the government ignores such consensus. Unfortunately, this is the case with the most significant attempt in the federal budget to increase the efficiency of government, through the “efficiency dividend”.

First introduced to federal public sector organisations in 1987 by the Hawke government, the efficiency dividend is a reduction of the budgets of these bodies by a certain percentage (usually 1.25%, but 2.5% in the current budget).

As the name suggests, this is intended to drive efficiency improvements. The idea is that public sector bodies do the same work with less resources and the government bottom line should benefit accordingly. Thus it is argued that this cut is not a cut at all, but merely a dividend from increased efficiency.

It is important to subject these justifications to scrutiny. As my research for the Centre for Policy Development points out, the efficiency dividend is the most significant initiative in May’s budget for driving more efficient government operations.

The measure’s predicted total saving of A$2.8 billion dwarfs the $530 million saved by the “Smaller and More Rational Government” initiative, which identifies a number of organisations for cessation or merger. This initiative is less of a blunt instrument in that it shows specifically what services will be affected, though the scattering of different organisations targeted makes it difficult to see any underlying rationale for the cuts.

An incomplete form of efficiency

The problem with the arguments for the efficiency dividend is that they take a very narrow view of efficiency. If the same results are obtained from fewer resources (this is questionable in some cases), this improves what is called technical efficiency. However, this ignores the “allocative” and “dynamic” aspects of efficiency.

Allocative efficiency is about ensuring resources are directed to the areas where they achieve the highest benefits. An across-the-board cut affects all public services regardless of the value they provide. This means allocative efficiency is not increased; worse, it may be reduced because the efficiency dividend’s effects are not even.

Smaller organisations and offices (such as the Australian Bureau of Statistics) feel a disproportionate level of pain, while larger organisations have more flexibility on where to make the savings. This means smaller offices serving regional Australia tend to be disproportionately affected. It also punishes more efficient organisations since they are expected to improve at the same rate as those with more numerous and easily addressed inefficiencies.

Dynamic efficiency, which involves adapting to change (including new technologies and modes of operating), is also damaged because the operation of the efficiency dividend is directly at odds with the dynamics of innovation.

First, it applies each year, yet innovations are “lumpy” with large opportunities in some years and less opportunity in others.

Second, innovations often lead to an apparent decrease in efficiency before the gains begin to show. For example, the introduction of a more efficient computer system will initially slow down work as staff learn to use it. The efficiency dividend takes away the resources first, meaning that the initial dip in efficiency occurs in a situation of constrained resources.

Third, working out innovative new ideas and ways to implement them often requires an investment of resources. The efficiency dividend ensures such investment is harder to find.

Experts unite against the efficiency dividend

A number of well-supported government reviews stressed the need to review the efficiency dividend. Examples include parliament’s joint committee of public accounts and audit’s 2008 inquiry and the 2010 Moran Review.

Tony Shepherd’s Commission of Audit criticised indiscriminate increases in the efficiency dividend. AAP/Lukas Coch

Cynics may reject these findings as self-serving, but they might find it harder to dismiss two more recent critiques. The National Commission of Audit, led by former Business Council of Australia head Tony Shepherd, was very critical of the common practice of governments to increase the efficiency dividend as a savings measure. The commission’s opinion is that cuts should be targeted with a clear rationale.

Even more damning was this year’s report by the Centre for Independent Studies (CIS), an organisation with a firm commitment to smaller government. The CIS recommended the “failed” efficiency dividend be abolished.

It is an indictment of any government’s commitment to efficiency that the most significant approach to driving more efficient operations is the use of such a blunt instrument, which flies in the face of condemnation from all sides. Complete disregard of advice on the efficiency dividend is unfortunately a bipartisan failing. It has survived through the governments of Hawke, Keating, Howard, Rudd, Gillard, Rudd again, and now Abbott.

The Rudd government’s economic statement in August 2013 increased the efficiency dividend to 2.25% for a period of three years despite all advice. The Abbott government has also gone against this advice, and the recommendation of its own Commission of Audit, by increasing the rate to 2.5%.

It is understandable, of course, that the Abbott government will not follow every recommendation. Governments must make decisions based on a range of different opinions, including from the departments of Treasury and Finance. Nevertheless, cultivating a more efficient government will require more rigour than the blunt and untargeted efficiency dividend. This approach is likely to be doing more harm than good.

Want to write?

Write an article and join a growing community of more than 139,900 academics and researchers from 4,249 institutions.

Register now