Using targets to improve public services (Institute for Government, June 2021)

Summary

Targets have been applied to a wide range of public services over the past 40 years. This report analyses whether targets improve the performance of public services and the reasons for this, making recommendations on when and how government should set targets. It focuses on national targets and examines what evidence there is for how they have affected how efficiently public money is turned into outcomes for the public.

Content

Key conclusions

• Targets have improved what was targeted. There is good evidence that the performance management regime in England, of which high-profile targets were an integral part, reduced hospital waiting times and improved exam results. Following the introduction of this system in the English NHS from 2000, the waiting times for elective procedures fell much faster than in Scotland, Wales and Northern Ireland, which had different performance management systems. In schools, the Welsh government’s decision to stop the publication of school league tables – which had effectively set a target for schools to outperform each other – led to students in Wales lagging behind those in England by the equivalent of almost two GCSE grades per student per year. There is also one good example of targets boosting overall performance and improving outcomes for service users. The four-hour A&E waiting time target resulted in a 14% reduction in the proportion of people dying within 30 days of attending A&E because of a reduction in waiting times for time-sensitive but hard to diagnose conditions, such as those suffering from strokes.
• Improvements in what was targeted were partly the result of gaming In hospitals, data was reclassified or manipulated, patients were left waiting in ambulances or on trolleys, and appointments that did not contribute towards meeting an explicit target were cancelled in pursuit of waiting time targets. In education, schools focused on pupils at the C/D grade boundary – at the expense of pupils far above or below this threshold – in response to the target for pupils to achieve five or more GCSEs at grades A*–C. Improvements may also have come at the expense of overall performance. For example, targets probably encouraged the government to provide disproportionate funding to hospitals at the expense of other, potentially more cost effective, approaches to improving population health, though it is hard to draw firm conclusions.
• Targets appear to be effective at raising minimum standards but not at driving excellence In health, the gradually decreasing elective waiting time target had most impact on the longest waits, and the scrapping of school league tables in Wales had no effect on the top quartile of Welsh schools, only the bottom three quarters, with pupils in the poorest performing and most disadvantaged schools seeing their exam performance suffer the most. Conversely, targets, by constraining staff, can create a culture of compliance that discourages innovation and prevents adequate services from excelling.