6.2 A high-performing public sector
6.1.1 Government savings and productivity
The Treasury's long term projections of the Government's capacity to supply goods and services underline the stark choices facing New Zealand and the vital role of productivity:
“…we simply can't keep doing what we have done without significantly increasing taxes or debt.” (The Treasury 2009)
The government's fiscal position deteriorated sharply from 2006, with substantial spending and tax decisions contributing to shift the Crown's operating balance from surplus to deficit, from plus $9.5 billion in 2006 to a projected minus $9.1 billion in the current year, and from saving to dissaving. At present, the government is borrowing some $300 million per week of longer-term debt (or $13.5 billion a year) - around 7% of GDP.[15]
The 2009 long-term fiscal projections were updated last year to reflect changes in the 2010 Budget. These produce lower debt-to-GDP levels in 2050 than previously forecast, but population ageing and other costs pressures still produce rising debt levels and an unsustainable fiscal long-term position. As new information comes in, the forecasts will continue to move about, with the effect that the 2050 debt-to-GDP levels forecast in Budget 2010 will be attained a few years earlier or later.
In the updated 2010 analysis, the base case projection, with government debt held to a sustainable level and public sector productivity growing at 0.3% annually, shows the supply of public services per capita in real terms declining in every year from 2013 to 2022, to about 8.5% below the current level, then slowly recovering to the current level in 2046.
However, if productivity growth is doubled to 0.6% per annum then the period of contraction of per capita services reduces from 33 years to 15 years; and by 2050 the annual level of services is 17% above the current level, compared with the base case's 5%.
Further testing of productivity options showed that if public sector productivity increased by 2% per year for five years and 1% per year thereafter, there would be no period of contraction, and growth in the supply of public services would be faster than the historic trend growth rate (Figure 6.1).
- Figure 6.1: Real public services per capita under different productivity scenarios

- Source: The Treasury
Alternatively, the same volume of public services as in the base case could be delivered and government spending reduced by $6 billion in 2020 and $25 billion in 2030. This would reduce net government debt by 33 percentage points of GDP in 2030.
Allowing public services to grow at the historic trend growth rate, with productivity growth at 0.3% a year and debt unconstrained, results in net core Crown debt at over 100% of GDP in 2050. And with productivity growth at 0.6% a year, net core Crown debt still increases from under 20% of GDP now to almost 60% in 2050 (Figure 6.2).
These debt profiles are clearly not viable for New Zealand and the protracted contraction in public services isn't really either. This is why raising public sector productivity is important, both in terms of its implications for the community, and in terms of options for increasing national saving.
- Figure 6.2: Net core crown debt with public services growth at historic trend (% of GDP)

- Source: The Treasury
It is important to emphasize that “productivity” and “improvement” are not simply camouflage labels for cost-cutting. There will be some, but the essence of the changes needed is more effective leadership, top-down and bottom-up initiatives to harness the thinking and ideas within an organization, supplemented by better practice, systems and technology from elsewhere, to produce more value, more efficiently. This is simply “business as usual” good practice in the private sector, and a common by-product when done well is a more empowered and enthused work force.
Box 3: Measuring government sector performance improvement
Interest around the world in measuring the performance of government-provided services has increased sharply given the GFC; constraints on government spending and debt; governments' large share of total expenditure in the economy; and the serious consequences of low productivity, poor targeting and waste in the public sector.
Productivity growth means growth in the volume of government services relative to the volume of resources used to produce those services, but it is only one measure of performance. Other performance measures include value-for-money and effectiveness.
Governments have commissioned reviews to find ways to measure outputs and outcomes of public services. These include the Atkinson review in the UK (Atkinson 2005), and the Stiglitz report in France (Stiglitz, Sen et al 2009). The OECD has been working on a manual on the measurement of non-market health care and education output. Services are usually much more difficult to quantify than goods and to incorporate measures of quality change. In addition there are usually no market prices for government services and this complicates the task of calculating relative values.
Last year, Statistics NZ released a feasibility study on measuring the productivity of public health services and education (Statistics New Zealand 2010b). In many countries, these are two services that incur the highest government expenditure. The study found that it would be possible to estimate productivity growth in these two sectors to near world-best practice and included a series of recommendations that would bring it up to the frontier.
It may be possible to apply these methods to other government services, as many of the issues and potential solutions apply more generally – however, Statistics NZ has not extended its research to data availability or suitability in these areas. The UK has studied the possibility of measuring the productivity of police, children's social care, social security, the criminal justice system, defence and adult social care.
Notes
- [15]This figure is gross borrowing and includes funding to re-finance existing debt and to pre-fund future planned spending. An annual deficit of $9.1 billion equates to an average of $175 million a week.
