Speakers:
Professor Julian Le Grand, LSE and Policy Strategy Adviser to the Prime Minister
Professor Anne Mills, Professor of Health Economics and Policy at the London School of Hygiene and Tropical Medicine

Chair:

Simon Maxwell, Director - ODI

1. The sixth meeting in the series was on the subject of 'Targets, Voice and Choice: Options for the Health Sector'. The speakers were Professor Julian Le Grand - LSE and Policy Strategy Advisor to the Prime Minister, and Professor Anne Mills - Professor of Health Economics and Policy at the London School of Hygiene and Tropical Medicine. The meeting was chaired by Simon Maxwell, Director of ODI.

2. Julian Le Grand focused mainly on choice issues and spoke to a PowerPoint presentation. He began by reminding the audience that there were different realms within which choice could operate: where (e.g. hospital), what (i.e. choice of treatment), when and how. The arguments varied according to what type of choice was being considered. He would focus on the question of where, and talk mainly about providers.

3. There were both intrinsic and instrumental reasons for focusing on choice. This was what people wanted, and increased choice could also provide an incentive to improve both quality and equity. However, there were also arguments against. It was often argued that people's main preference was for strong local services. It was also argued that effective choice required excess capacity in the system, that the poor would be disadvantaged, and that institutions would pick easy cases ('cream-skimming').

4. It was important to begin by noting that the evidence was that people did want more choice and control (though the elderly less than others). Polls showed that the poor (who in fact had the least choice) wanted more choice more than the rich. There were also currently large inequities in the system already, which needed to be taken into account.

5. The real reason for considering greater choice was that it would provide incentives to improve quality. A monopoly service would only work if professionals were highly motivated and there was a high degree of trust (workers were 'knights not knaves'). Furthermore, it was necessary that voice mechanisms should be effective and that centrally driven performance management should deliver results. These conditions were hard to meet. Workers in the health system had complex motivations, mechanisms for voice were 'clunky' and tended to favour the middle class, and targets had led to long-term demoralisation of the work force.

6. The answer, then, was to extend choice, with money following choices. UK experiments had shown that choice instruments (for example, choosing alternative hospitals) had led to faster treatment. However, there were some lessons: there needed to be alternatives (for example, the availability of unused beds), there needed to be good information, transactions costs needed to be kept low, and users not providers needed to make the choices, so as to avoid cream-skimming. It was important that the system should be sensitive to finance, for example by using administered prices.

7. Julian Le Grand's conclusion was that choice mechanisms did have great potential. But systems needed to be well-designed, and it would probably be necessary to provide special assistance to help the poor navigate a new system (for example through Patient Care Advisors). For the longer term, there were also some options to consider to reduce cream-skimming, for example stop-loss insurance and risk adjusted tariffs.

8. Anne Mills also spoke to a PowerPoint presentation. She began by reminding the audience that in developing countries the non-State sector was extremely important. For example, in Tanzania, 65% of patients seeking treatment for malaria used shops rather than formal medical facilities. This was because access and opening times were better, waiting times were lower, and staff were less rude. On the other hand, diagnosis and advice were sometimes inadequate, and drugs were sometimes of poor quality.

9. Reviewing the various options for improving health service delivery, Anne Mills noted that targets were indeed problematic, with a severe risk of perverse incentives and de-motivation of staff. Voice options were currently very weak in most developing countries. Choice, on the other hand, already existed in many situations. Critical questions were whether people could be helped to exercise choice more effectively (for example by providing vouchers for bed nets).

10. On the other hand, it was very hard to provide choice for treatment options. The system generally had low capacity, and the problems identified by Julian Le Grand in the UK were multiplied in developing countries. An interesting example of contracting out was cited from Cambodia, where large international and national NGOs had contracted to provide health services in districts. This had led both to better and more pro-poor coverage.

11. Anne Mills' conclusion was that the health system was not homogenous and that it was dangerous to generalise. It was easier to offer choice options for simple rather than complex services.

12. A number of issues were raised in the discussion: some participants thought that the priority should be to provide a well-functioning, State-run system for health care. Others thought that there were significant advantages to opening up the market, for example through contracting or other choice mechanisms.

13. At the same time, choice options were highly complex. For example, the administered price regime proposed for the UK would be difficult to administer. At the same time, the diversity of options would be extremely complex for patients.

14. The Cambodia example offered an interesting possibility, in which competition was introduced at a relatively high level, with the decision criteria resting on the reputation and overall capacity of the NGOs to provide a service.

15. In general, there was some enthusiasm for greater exploration of public-private partnerships in health provision.

Simon Maxwell

30 November 2004