Click here for search results

Leadership Forum (February 2004)

Author: Davidson R Gwatkin, served until March 2003 as the World Bank's Principal Health and Poverty Specialist. He is now a consultant on health and poverty to the World Bank, the Rockefeller Foundation, and other organizations.

Documenting and lamenting health inequities is much easier than doing something to reduce them. Perhaps this helps explain why so many people concerned with health equity are currently suffering from a “threshold paralysis” that impedes a shift in focus from documenting inequalities toward identifying ways to reduce them by reaching poor groups more effectively.

What can be done to overcome this paralysis? And what strategic issues are likely to arise once it is overcome? Those are the two questions with which this note will deal.

Taking the Next Steps

The journey toward the discovery of truly effective ways to reach the poor may well prove to be long, along paths that remain to be discovered. But as the famous saying puts it, a journey of a thousand miles begins with a single step. And two initial steps in the journey toward equitable health care are readily visible.

The first step is to learn from the experience of what has been tried before. This means reviewing, from a distributional perspective, the experience of prior attempts to reach the poor; and then applying the lessons learned from those reviews to the design of new programs.

Such distributional reviews have rarely been done. Instead, examinations have usually been limited to looking at the record of interventions in an entire population, rather than at how well the interventions reached and benefited disadvantaged groups within the population.

One effort to fill this gap, by learning from and encouraging further distribution-oriented reviews of recent experience, is the World Bank’s Reaching the Poor Program. The Program began by commissioning studies covering over 35 health, nutrition, population interventions in 15 countries. The studies, selected by professional peer review from among the 150 proposals received in response to an internationally-distributed announcement, feature the use of incidence analysis to determine how the interventions’ benefits are spread across socio-economic classes within the societies they covered.

These studies, now largely complete, will be among those presented at a global conference to be held in Washington February 18-20, titled “Reaching the Poor: What Works, What Doesn’t and Why?” At the conference, the 250 participants will review the evidence presented, much of it quite promising, and seek to develop guidelines for designing programs more effectively oriented toward the disadvantaged. If all goes well, the studies and conference discussions will both provide guidance to program managers, and also inspire other investigators to take up incidence research.

The second, equally important step is to experiment in the field with better ways of reaching poor groups, and to monitor the results. Up to now, such field experimentation has focused on increasing coverage in the overall population served, without regard to how the increase has been distributed across better-off and disadvantaged groups. As a result, the experiments provide little guidance for righting the distributional imbalance that typically occurs.

An illustration of what’s needed to fill this shortcoming is the first known instance of distribution-oriented field experimentation. The experimentation is being organized by the INDEPTH Network, an association of 34 surveillance sites covering populations of up to 200,000 in each of 18 countries. An earlier diagnostic exercise, based on existing data from 13 INDEPTH sites, had shown the existence of significant health status and service coverage inequalities in what had previously seemed to be homogeneous poor populations. That finding gave rise to field research with approaches designed to reduce some of the service coverage inequalities identified. To begin, INDEPTH is supporting projects in three sites, in Ghana, South Africa, and Tanzania. The three projects seek to increase knowledge about AIDS and the use of AIDS services by disadvantaged population groups, through increased visits by community workers to poor households.

These projects are particularly welcome in the area of AIDS, which has thus far seen even less poverty-oriented work than other areas of health service delivery. But they are even more important as illustrations of the kind of work that’s also badly needed in other areas.


Facing the Next Issues

Both the steps just suggested involve special efforts to reach out and serve disadvantaged groups – that is, looking to targeted programs that give special priority to the poor in order to achieve equal coverage. Is equal coverage of services the best strategic objective to adopt? Not everyone thinks so. Many would argue instead for working toward universal coverage – seeking to cover everyone, rich and poor alike.

The choice between equal and universal strategies appears to be emerging as a central issue attracting the attention of people concerned with reaching the poor. There are important points in favor and against each option.

Universal coverage, if achieved, would fully cover the poor as well as better-off, thereby attaining complete access to services for all the poor. Moreover, it is inherently egalitarian, in that both poor and rich would be not only covered to the same degree, but also by services of equal quality. In addition, a universal coverage strategy can be easier to initiate, because of its potentially stronger appeal to politically-important upper-income groups dubious about preferential treatment for a particular group to which they don’t belong.

There are also significant limitations to a universal coverage strategy. Perhaps the most important is the likelihood of at least a temporary increase in poor-rich coverage differences. In the many situations where significant numbers of better-off people remain uncovered, the well-established “inverse care law” (that access to care and the need for it are inversely related) suggests that these better-off would be the first to take advantage of any expansion of services, with the poor gaining only after the needs of the better-off have been fully met. Also, attaining universal coverage of all services would usually be prohibitively expensive, so that the most that can reasonably be hoped for is probably attainment of that objective for only a limited set of services.

Effectively targeted services aiming to attain equal coverage would overcome the principal limitation of the universal coverage strategy just mentioned, by bringing benefits to the poor right away and thereby immediately beginning to reduce poor-rich differentials. Thus, the poor would not have to wait.

Targeted services also have important limitations. One is the stigma often associated being deemed poor enough to qualify for targeted services. A second is the frequently lower quality of services delivered through targeted programs: “services for the poor are poor services,” as the old adage claims. Further, targeted programs are notoriously difficult to operate, as targeting remains an inexact science that requires careful and sustained effort.

So which strategy should be followed? A quest for universal coverage through programs that reach everyone, poor and non-poor alike? Or the pursuit of equal coverage through targeted programs oriented especially toward the poor? Here’s a suggestion:

  • There are at least some situations where a focus on universal coverage and/or untargeted programs would be the most sensible strategy. One such situation would be a setting where nearly all upper-income people already have access to a particular service that is largely unavailable to the poor. In a setting like this, the poor would of necessity be the principal beneficiaries at all stages of progress toward universal coverage. Another case calling for untargeted programs would be that of a communicable disease whose transmission can be broken at a relatively low overall coverage level. Here, the most reasonable strategy would be to cover those who are easiest to reach and cover, even though those people may well be the better off, since such limited coverage would be sufficient to protect the poor as well as the non-poor. A third, important instance would be a country with a government adequately committed and competent to ensure that universal coverage is fully achieved within a reasonably short period of time.
  • In most cases, the poor would be better served through targeted programs that seek to achieve equal coverage. The justification for this view lies not in the strengths of targeted approaches, but in a central weakness of a universal coverage strategy: the limited likelihood of reaching the strategy’s objective. Governments committed to and capable of fully achieving universal coverage are few and far between. Elsewhere, a much more likely scenario for a universal coverage approach would be initial enthusiasm that wanes well before reaching the approach’s admirable but rarely attainable goal. This would produce large gains for the better-off who would be the primary beneficiaries of the approach during its initial phase, as noted earlier; but only limited improvement among the poor, for whom major benefits would depend on the approach’s being fully carried through. Thus, while universal coverage strategies are unquestionably very attractive in principle, in practice they are much likely to bring only limited benefits to the poor, and to produce increased differences between poor and rich.

For additional information on the results of this survey, please contact Davidson R Gwatkin

Further Reading

Conference Information

Permanent URL for this page: