Human resources impact assessment.

For decades, discussions on human resources in health have ended with a ritual call for more and better manpower planning. But this traditional wisdom has been discredited by unrealistic or vague targets, based all too often on information that was inaccurate, outdated and unrelated to the policy agenda. Nevertheless, in as labourintensive a domain as delivering health care, reform does entail far-reaching adjustments in the workforce and a new definition of the roles of health workers (see Buchan & Dal Poz’s article on pp. 575–580 of this issue).Meanwhile, globalization adds impetus to the migration of health workers with its destabilizing effect on health care delivery (1). Many decision-makers readily point to human resource problems as the chief bottleneck they face in attempting to scale up health systems. Yet time and again the reform agenda neatly skirts around the sensitive and difficult issues involved—not least because there aremajor gaps in the knowledge base required for a realistic workforce strategy. Today’s emphasis is no longer on the mechanics of optimizing the quantities, skills and distribution of manpower. The ‘new’ concerns claiming most attention in discussions are: the implications of the public/private debate; decentralization and civil service reform; performance management; and staff retention. In health workforce planning in most countries, training and employment used to be regarded as an essentially public sector affair. Nowadays, however, it is hard to imagine a debate on human resources that does not refer to the private sector. Partly because we have more documentation on it now, we talk more readily about what was still a taboo subject in the 1990s: the fact that throughout the world, public sector health staff boost their grossly inadequate incomes with private practice, often in an ambiguous context that compromises their public responsibilities (see Ferrinho et al. on pp. 581–584 of this issue). On the whole, however, the wealth of opinions on health workers in the private sector contrasts strikingly with the lack of empirical information, particularly in developing countries. Response to human resource problems — particularly those related to income and performance — is often piecemeal and improvised. Few countries propose structural responses other than decentralization. There have been situations in which the greater managerial freedom made possible by decentralization has helped to improve things. On the whole, however, there is no evidence for an automatic link between decentralization and more effective management of human resources or greater efficiency. Health workers themselves tend to be the most sceptical about decentralization. It can reduce job security and limit upward career mobility. It brings in the destabilizing prospect of being hired, disciplined or fired by local authorities or committees which are less predictable than the national ones. More often than not, it has met with stiff resistance among health workers (2). As decentralization brings human resources management closer to the actual operations, increased client pressure pushes managers towards performance management. The principle is deceptively simple: explicit objectives and targets steer individual performance, linking it to broader service and organizational goals. Once performance in relation to the set targets is measured it becomes possible to promote desired behaviour through financial or other incentives and disincentives. The drive towards formal performance management has certainly improved the information base for human resource planning and helped to institutionalize continuous medical education. But there is very little evidence that formal performance management systems actually affect quality or patient outcomes, and none to show that any gains in efficiency outweigh the costs of setting up the systems. To be fair, there is no evidence to the contrary either, but it would be naive to look upon performance management as the magic solution for problems that planning failed to solve. What then can one do? The challenge is not unlike that of getting the environment onto the development agenda a decade ago. Like the environment then, human resources for health now are recognized as a major problem. Like concerns about the environment until recently, these are rarely, if at all, translated into policy interventions. The problems are similarly complex, context-dependent, and often unexpectedly made worse by well-intentioned projects or reforms. Environmental concerns are now more systematically taken into account partly because decision-makers started to ask as amatter of course for explicit environmental impact assessments whenever major development plans came up for approval. Very often, such exercises were no more than an administrative formality, but as a whole they played no small part in getting environmental concerns into the mainstream of policy-making. Policy-makers and donors concerned with human resources problems may want to go down a similar road. They may request those proposing a major new project or policy to make a systematic and formal ‘human resource impact assessment’ during its preparation. Such assessments would examine the likely effects of the proposed project or policy on the health workforce. This would yield a triple benefit. First, it would draw the attention of decision-makers to the potential consequences of their decisions for the human resources in their health system. Second, it would help steer organizational and financing decisions towards minimizing negative effects on the workforce and enhancing positive ones. And third, it would help to build up documentation on how human resources are affected by new policy initiatives — information that is sorely lacking at present. Major initiatives such as the Global Fund to Fight AIDS, Tuberculosis and Malaria could only benefit from asking applicants to assess the likely consequences of their proposal on the situation of human resources. That would be no substitute for a global workforce strategy, but it would be a start. n