How can we better measure public sector productivity?

23 Feb 17

New Zealand has recently started looking into the thorny issue of how to measure and improve productivity in the public sector

On one level Kiwis should be pleased with how New Zealand economy is going. The public finances are healthy, there has been solid growth in GDP, rates of labour utilisation are high and major policy settings are stable. Yet New Zealand had the fourth lowest labour productivity growth of OECD countries between 1995 and 2014. And given population ageing (slowing labour market growth) future prosperity will to a larger degree depend on increases in labour productivity.

Similar issues face public services too. New Zealand’s public services often rank highly in international comparisons. Yet, as Statistics New Zealand has estimated, between 1996 and 2014 increases in outputs of the public sector have largely been driven by increasing inputs. In health, average growth in output of 4.0% reflected input growth of 2.9%, with labour productivity contributing 1.1%. In education, output growth of 1.1% reflected input growth of 2.5% and labour productivity falling by 1.4%.

Measuring the productivity of public services is complex and the Statistics New Zealand data do not account for things like changes in quality. Yet when seen in the context of the Treasury’s work on the long-term fiscal outlook they still provide food for thought. Not only will the demand for key public services increase but we can also expect growth in the aggregate labour force to slow. The implication is that public sector managers can expect their operating environments to become both more demanding and more input constrained, and so they are going to need to increasingly focus on lifting public sector productivity.

But this is a good thing. Consider fiscal policy. A greater growth rate of public sector productivity would allow more public services to be provided at the same cost – or alternatively the same level of public services to be provided at a lower cost. Faster public sector productivity growth can bend down fiscal cost curves and provide future governments with fiscal headroom. More productive public services would also boost the economy's aggregate productivity growth given the share of the economy for which the public sector accounts.

And the effect of public services goes well beyond these direct measures. They also have indirect effects, as productivity growth in the private sector depends on a healthy, well-educated population, whose efforts are enhanced by physical and social infrastructure. Given their importance it is reasonable to want to maximise the service provided for a certain level of inputs. A concern with productivity is a feature of strong public services.

But while greater public sector productivity is a good thing it is not immediately clear what role productivity measures should play in the performance management of these services.

It is in this context that the Productivity Commission and Professor Norman Gemmell, the Victoria University of Wellington Chair in Public Finance, have been investigating measures of public sector productivity. Better measurement is the first step in lifting productivity. This is also a developing field internationally and work is needed to develop metrics that provide practical insights for public sector managers.

A stock response to the idea of measuring public sector productivity is that it is too hard or that there is something unique about public services that make this impossible. But this is an old view. There have been years of work by national statisticians and others developing techniques for measuring public sector productivity. Important lessons have been learned.

Conceptually a key problem is the lack of market-clearing prices in the public sector. To explain why this matters it can be useful to think about the foundations of productivity measurement. Given the diversity of outputs produced in an economy, productivity measures require an approach for combining diverse outputs into a single index. Take a simple example of an economy that produces only guns and butter. Estimating productivity requires a measure that combines the output of both these products. But just how many kilos of butter are equivalent to one gun?

In the private sector, prices can be used to make these comparisons. Different outputs can be combined into a single index based on their value in the marketplace. This approach is taken because prices are assumed to be a good indicator of consumers’ valuation of (willingness to pay for) different outputs.

In contrast, public services typically lack, or at best poorly reflect, prices as they are provided free or are subsidised. Hence prices cannot be used as proxies for the value the services generate and some other way is required to combine diverse outputs into a single index (a weighting schedule). Quite often cost weights are used but these reflect the value placed on the service by the producer and imply that higher costs equal higher quality.

And the challenges do not stop there. The quality of a product or service may vary over time as it is refined and developed or as the producers’ operating environment changes. Yet adjusting estimates of public sector productivity for quality changes is a complex exercise. Consider how the United Kingdom Office for National Statistics had to revise their approach to quality adjusting education quantity.

And this is all before we account for institutional factors (e.g., Parliamentary accountability for inputs, roles of competition and consumer choice, etc.) that may mean public services may require a different approach to measuring productivity from that used for the private sector.

But while this may be difficult it does not mean giving up. Given the importance of public services we cannot ignore the need to measure and improve their productivity. For some services and tasks this will be easier than for others, and the best approach will be tailored to the users of the data and the questions to which they need answers. Done properly efforts along these lines will support public sector productivity growth and stronger public services.

  • Patrick Nolan

    Principal advisor, economics and research, New Zealand Productivity Commission
     

Did you enjoy this article?

Related articles

Have your say

Newsletter

CIPFA latest

Most popular

Most commented

Events & webinars