LSE research has been instrumental in providing guidance as to how to do “productivity path analysis” in different types of central government agencies.
What was the problem?
In advanced states, the government sector accounts for 25 to 35 per cent of all final consumption of outputs. But for many decades now its productivity has gone unstudied.
The essential idea of productivity is to divide an organisation’s total outputs by its total inputs, yielding a metric whose over-time development can be compared. The concept has been fundamental for firms (and governments) in developing knowledge of private sector growth processes. Public sector productivity has previously been seen as too difficult to measure, especially at department/agency level. While total inputs are now well measured within government, attributing inputs to outputs is somewhat harder.
The real core difficulty has been in specifying a measure of “total outputs” for a public agency. State agency outputs are diverse and not (in general) marketed, and so they cannot be effectively priced. Without prices, different types of outputs could not in the past be weighted against each other. Hence the convention became (in national statistics and other sources) that government outputs were always costed as the same as total inputs, meaning that government productivity was assumed to be flat over time.
A breakthrough study by Atkinson in 2005 solved this issue by proposing that public sector outputs should be weighted by their administrative costs, so as to yield a cost-adjusted total output. This idea was for a time used in an experimental way in several countries’ national statistics, where it ran into some problems. But the impact on government agencies was small at this stage, since productivity calculations were made only at the whole-nation or national sector level.
In 2013, Professor Dunleavy and Dr Leandro Carrera published the first systematic book on government productivity, Growing the Productivity of Government Sector Services, showing how applying Atkinson’s approach can produce a “productivity path analysis” for each central government department and agency, that tracks over time their cost-weighted total outputs divided by total inputs.
The study demonstrated the immense impact of information and communication technologies (ICTs) in strongly growing the productivity of some UK public agencies (like taxing and customs departments). This contrasted with other departments (like the Department for Work and Pensions in social security), which had avoided digital changes and so had not improved their productivity at all over two decades.
What did we do?
Following on from the favourable reception of Professor Dunleavy and Dr Carrera’s book, Growing the Productivity of Government Sector Services, in 2016 the Organisation for Economic Co-operation and Development (OECD) focused its annual theme on productivity and invited Dunleavy to give a keynote speech to the Annual Meeting of Budget Officers - an audience of around 120 senior Treasury and finance departments from 35 OECD countries.
Dunleavy’s paper focused on how managers could apply methods from his book, and provided for the first time a comprehensive guide on how to do “productivity path analysis” in different types of central government agencies:
- Caseload agencies (with easily trackable units of output).
- Agencies with complex or “lumpy” outputs.
- Regulatory bodies.
- “Policy only” agencies.
His presentation was well received, and a revised version of the paper was published in the OECD Journal on Budgeting, entitled “Public sector productivity: measurement challenges, performance information and prospects for improvement”.
What happened?
Following on from work on digital change in the federal government, the Australian Department of Finance commissioned Professor Dunleavy and Professor Mark Evans to develop improved measures of organisational effectiveness that might allow them to implement greater budgetary flexibility for departments, while fostering efficiency gains and service delivery improvements. Their first report recommended the development of “productivity path analysis” (PPA), so that agencies demonstrating trend levels of productivity improvement or better could be assigned greater budgetary freedoms. A co-production method for developing the detailed application of these proposals was recommended.
The Department of Finance commissioned a second stage report, and Evans held a set of co-production workshops with senior officials representing all departments focusing on the PPA template. These 2017 sessions largely ironed out most of the difficulties in measuring and implementing PPA across agencies and made advances in winning acceptance of productivity measurement, and in specifying how it can be conducted.
In mid-2017, the New Zealand Productivity Commission (NZPC) approached Dunleavy to advise them on a forthcoming report on public sector productivity. Since New Zealand is a unitary state, NZPC has the potential to affect service measurement and improvements across all levels of government. In the July, Dunleavy met with the Commission Chair and Chief Economist in London and explained the approach. He then consulted on the substantial NZPC draft report.
In spring 2018, the New Zealand Productivity Commission (NZPC) asked Dunleavy to visit New Zealand and address their main consultation conference in Wellington, which attracted 250 public sector and consultant attendees. He also gave four detailed briefings on productivity measurement to central departments and local health authorities and discussed with NZPC board members the progress of their report. In August 2018, the Commission subsequently published two “state of the art” reports in the government sector, both referencing Dunleavy’s work.
Two main difficulties remain for developing government productivity measures and implementation. First, despite progress on applied measurement, the Australian study process (involving co-production seminars) demonstrated that some departments still feared that they would be disadvantaged in budget terms by the introduction of productivity measurement. And second, central government budget regulation generally requires all departments (and hence ministers) to operate in a standardised set of rules. APS officials and ministers still see this as vital in retaining exactly equal treatment of departments.