A common refrain these days amongst punters and politicians alike is Australia’s productivity slump – even prior to the government’s announcement of the August 2025 “Economic Reform Roundtable” with a strong focus on lifting productivity growth. Weak productivity growth is seen, correctly, as a headwind to improving living standards.  Any threat to our future standard of living provokes scapegoating, with fingers variously pointed at workers, managers, government and even our climate.

So, with the roundtable imminent – and given its focus on both productivity and budget sustainability – it’s worth (re)asking some fundamental questions. What really is “productivity”? How is it measured? What generates (and holds back) productivity growth? And, critically, how can we meaningfully gauge productivity in the non-market sectors that take up more and more of the budget?

What is productivity?

Productivity can be conceptualised as the value produced by a production process, divided by the value of the set of inputs used in that process. Starting with this simple ratio helps us understand the many ways that productivity can be increased or decreased, as well as the inherent perils in attempting to measure it.

Start with the numerator of the ratio. The idea of “value produced” may sound intuitive, but it is actually quite slippery. How does one come up with a number that captures the value of the service received at a café or a bar, or the enjoyment of night-time runners using a park whose lighting has been upgraded?

In neither case is the value produced and enjoyed by the consumer directly priced by a market mechanism. In a sufficiently competitive market that has achieved what economists would call a stable “long-run equilibrium”, service quality at a bar or café would be reflected in the price of what you pay for your experience. However, disentangling the value of the service per se from the value of the food or beverages bought at wholesale, prepared and served is a further complication.

On top of this, dynamism is ever-present, so the relationship between service quality and price is shaky. Not only are competitive pressures variable, but “equilibrium” is mainly a figment of the economist’s imagination, because everywhere, all the time, people are learning and changing and therefore markets are changing, and hence operating “out of equilibrium”.

Moreover, some types of value never get priced, even in “perfect” conditions (think of the night-time runners). Moreover, in markets that are subsidised or otherwise not freely competitive, price and value may systematically diverge.

In cases where the measurement of output value is problematic, it can be tempting to simply use the denominator of the ratio (i.e., input value) as a substitute. Mechanically, however, this erases productivity. Furthermore, measuring input value is still not simple: the price paid for some number of hours worked, for example, is fraught because different workers bring different skill sets, experience and interests, and so an hour worked by two different workers surely does not contribute the same value into a firm’s production process.

“Capital” (e.g., machines) also comes in different varieties, and the purchase price of a machine is not a perfect proxy for its value, especially given constant technical progress.

What gets measured?

With all that as background, spare a thought for the head of the Australian Bureau of Statistics (ABS), who must generate myriad economic statistics that can be defended as capturing the state and strength of Australia’s economy – amongst them, measures of productivity.

In the private or “market” sector, the method the ABS uses to measure what is termed “multi-factor productivity” (MFP, i.e., productivity from all inputs) is based on trying to measure both the numerator and the denominator of the core ratio, and arriving at a measure of the value added by Australia’s production processes that turn inputs into outputs. Sector-specific estimates (and estimates of only the labour component of inputs) are also constructed.

However, over 25 per cent of Australia’s economy is represented by the public sector, which is largely shielded from conventional competitive pressures and whose output often carries no price. For such sectors, input prices (production costs) are generally used as proxies for production values, and MFP estimates are simply not published. However, for health and education, attempts are being made to construct productivity estimates based on output volumes (e.g., the number of emergency-room presentations or graduates) rather than output values.

Where do productivity gains come from?

Imagine that every production process in the country was using the most state-of-the-art protocols and equipment, such that productivity was as high as technologically achievable today. How, then, could productivity grow to a sustainably higher level from this year to next year?

The only way to increase productivity beyond the present technological frontier is to push out that frontier. The only ways to push out that frontier are to create new markets (i.e., to discover new things to make that people truly value) or to find better – i.e., more efficient – means of doing existing things. There are many ways to find these new things and better means, but all of them involve creativity and costly experimentation.

If some productive processes are not already at the technological frontier, then one can increase productivity in the short run by getting them to that frontier via the dissemination of technology. This opportunity exists in many less-developed nations today, and to a lesser extent in developed countries such as Australia.

Ultimately then, sustainable productivity growth requires an environment that accommodates and rewards creativity and experimentation.

Where do (apparent) losses in productivity come from?

If the value of labour as an input to production is captured by its wage, as is often assumed, then ever-increasing wage levels exert a persistent drag on measured productivity growth – though one that most people benefit from.

Another drag on productivity growth is the time that it takes employees to ensure that a business complies with the regulations applicable to its size and industry. The ideal regulation is one whose benefit – in terms of safety, competitiveness or quality and hence ultimately output value – exceeds the costs that must be paid to comply with it.

However, regulators have little incentive to submit their proposals to an independent, whole-of-economy cost-benefit test given their narrow mandates. Ironically, regulators themselves fall into the “public administration and safety” sector for which the ABS publishes no measure of productivity.

The result is a regulatory burden on companies that produces unknown output value while being an always-on vacuum for input value, pushing down productivity.

A second drag on measured productivity (if not necessarily actual productivity) is a gradual shift of Australia’s industrial mix away from sectors where productivity is easily measured and toward sectors where it is not – such as the caring professions and public service.

Non-market productivity measurement, and other conundrums

Non-market activities, such as volunteering or unpaid caring for others, are massive inputs to the economy but are not priced, and their productivity is unmeasured. Similarly, as Australia hires more public servants, we reduce the pool of workers in jobs for which measuring productivity is straightforward. So, there is a real premium on identifying creative ways to think about public sector and other non-market productivity.

For example, measuring the productivity of childcare (itself a potential input to productivity) is nearly impossible, since it is co-mingled with other inputs, such as food and parental time, and the quality of the main “output” aimed for (a well-functioning, happy person) is revealed over a lifetime following the initial care investment. The closest the ABS gets to a productivity measure in this sector is to use input costs as a proxy for value, again mechanically erasing productivity.

Measuring the productivity of public sector work beyond care services through metrics like customer satisfaction or project completion may sound useful until one considers applying the same metrics to a private firm. Are customer satisfaction rates really a reasonable proxy for the market value of whatever a firm makes?

Even if one accepts imperfect measures of productivity, aggregate productivity calculations can mask problems. For example, the performance of universities that hire both administrators and academics is ascribed to all employees. This hides the lower or even negative contribution to productivity of excessive numbers of administrators.

Where to from here?

Given the significant difficulties in direct measurement of productivity for many sectors, including those experiencing disproportionate expansion, one approach is to reduce our efforts to innovate in this direction and instead to find a complementary way to try to assess productivity.

For instance, the Productivity Commission recently suggested a new way of measuring productivity in healthcare that uses quality-adjusted life years (QALYs), a plausible measure of health quality, rather than output volume measures.

More fundamentally, though, a sustainably productive economy is, logically, one where most people work most of the time on value-creating activities that they were trained for, or on activities that are good for them and their communities’ long-run health and development.

Following this logic, we could examine more closely how people spend their time. We can infer much that is relevant to productivity growth from observing, for instance, that business owners are spending more time filling out compliance paperwork, or that those discharged from hospitals are spending less time recuperating, or that children are spending more time in front of screens.

Australia’s last “time use” survey ran in 2006. The Economic Society of Australia ran a community education event recently to emphasise the policy relevance of this survey.

To support Australia’s structural productivity growth, we should systematically find and reduce drags on productivity (excepting steadily rising wages), and promote an environment where experimentation is valued and rewarded. Taking a fresh approach to inferring productivity – especially of non-market activities – using time use data rather than dubious direct measures can help focus these efforts.

Gigi Foster is Professor at the UNSW School of Economics and Senior Scholar at the Brownstone Institute. She works in diverse fields including education, social influence, time use, lab experiments, behavioural economics and Australian policy. A nationally awarded university educator, Gigi was named 2019 Young Economist of the Year by the Economic Society of Australia and co-founded Australians for Science and Freedom in 2023.

Image credit: Roberthyrons from Getty Images