Training world-class researchers is essential to the development of a powerful science base for the European knowledge economy, and must therefore form a major part of European research and development (R&D) policy in the coming years. But a top-down regulatory framework for doctoral training that might hinder rather than help universities to compete globally should be avoided, according to the latest report from the League of European Research Universities (LERU). 'Frontier research in all disciplines is a major source of innovation in a knowledge-based society and therefore has a crucial role to play in the development of European economic, political and cultural life. Highly trained researchers, and especially graduates of PhD programmes and other research-based doctorates, act as major agents of creativity, innovation, knowledge transfer and problem solving,' writes LERU in its report 'Doctoral Studies in Europe: Excellence in Researcher Training'. LERU believes that to drive forward research and, in turn, the knowledge-based economy, doctoral training in Europe must focus on excellence. This pursuit of excellence requires a strong selection and evaluation of candidates, their immersion in a fertile and rich research environment, high quality administrative and programmatic support, and strong institutional leadership and management. But in order to achieve this level of excellence, Europe must first address the fragmentation of its doctoral training programmes. There are currently over 1,000 universities conferring doctoral degrees in Europe. In contrast, there are only 400 PhD-awarding institutions in the US. Less than 100 of these are responsible for awarding 80% of all PhDs. LERU points out that because of this fragmentation, energy and funds are being dispersed, resulting in a diffused impact. It argues that to remain competitive internationally, high level research and training must become more concentrated and focused. One suggestion is for better cooperation between institutions and the introduction of PhD programme network initiatives. But greater concentration does not necessarily mean less choice. Universities are finding it increasingly difficult to fulfil the varied roles expected of them. These range from a learning and research provider to an innovation incubator. 'Europe would therefore benefit if a broader spectrum of institutions were to emerge, ranging from globally competitive, leading edge, research-intensive universities operating at the frontiers of research and education, to those most deeply engaged with their local communities in satisfying the local demand for graduate skills, training programmes and market-driven consultancy and advice,' writes LERU. At their meeting in May, EU Ministers responsible for Higher Education called for doctoral programmes to be embedded in institutional strategies and policies, and for the development of appropriate career paths and opportunities for doctoral candidates and early stage researchers. While broadly agreeing with the ministers' statements, LERU warns against the prospect of a top-down regulatory framework for doctoral training, which it says might hinder rather than help universities compete globally. Credit or accreditation systems for doctoral training should be avoided, while definitions of study fields, in duration, in organisational structures, and in admission criteria, must be decided by institutions, says LERU. 'Regulation of doctoral training must be kept to a minimum. European universities should not be handicapped in their global competition for the most talented doctoral students by bureaucratic and restrictive regulations which their principal competitors do not have to face. The principle of institutional autonomy means that universities must be free to develop their own strategies, methods, commitments, and organisation of doctoral education,' argue the authors of the report. Instead, LERU recommends that governments embed quality assurance of doctoral training in the regular assessment of research degree-awarding institutions. Quality assurance should be based on the quality of supervision, high completion and reasonable time-to degree rates, which are the essential parameters of success for a PhD programme, rather than bureaucratic monitoring, notes LERU.