In This Issue


  • Bradford H. Gray

A great deal of evidence from many countries links higher income to better health status. Even so, the nature of that association still remains controversial. Although higher incomes could be leading to better health, it also could enable people in better health to earn higher incomes; that is, income and health could have shared causes. Whether increases in income make a difference in people's future health status is uncertain as well, and this is a particularly important policy question regarding people with low incomes. The first article in this issue of The Milbank Quarterly addresses this question. In “Upstream Solutions: Does the Supplemental Security Income Program Reduce Disability in the Elderly?” Pamela Herd, Robert Schoeni, and James House review the published evidence regarding whether socio-economic status is a “fundamental cause” of differences in health status (Link and Phelan 1995), and they present new data on this question.

Fundamental cause theory posits that socioeconomic status is not just a proxy for other factors that may influence health, such as access to health care, sedentary behavior, smoking, and obesity. One body of evidence supporting this claim shows that neither changes in these factors during the twentieth century nor shifts in the major causes of mortality from infectious to chronic diseases altered the link between low socioeconomic status and poor health. If low income is in fact a fundamental cause of health problems, then population health policy should place more emphasis on factors like social and economic status. Herd, Schoeni, and House examine whether the Supplementary Security Income (SSI) program, whose purpose is to raise the income of the poorest elderly Americans, affects that population's level of disability. The authors use within-state changes in SSI benefit levels between 1990 and 2000 to examine the effects of these changes on the states' rates of disability. Their results support the hypothesis that more generous SSI benefit levels lead to reductions in disability in the poor elderly. For example, an increase of $100 per month in the state's maximum SSI benefit for single elderly individuals led to a decrease of almost a half percent in their rate of mobility limitations. Since only about 10 percent of single elderly individuals receive SSI benefits, the effect for that population could be up to ten times larger.

The United States spends much more on health care than do other industrialized countries yet lags behind them on major health indicators. Although more research is needed, the important point is that income support policies may provide another way to resolve this dilemma.

The second article in this issue also is concerned with disability among the elderly in the United States. Like the first article, it provides both an extensive literature review and new empirical evidence. Survey data show that the rate of disability for the elderly has been declining since 1980. In “Why Is Late-Life Disability Declining?” Robert Schoeni, Vicki Freedman, and Linda Martin provide the most comprehensive examination, to date, of the reasons for this phenomenon, including new analyses of data collected from the National Health Interview Survey between 1982 and 2005. Many factors seem to be involved, including a greater use of assistive technology and the declining rates of heart and circulatory conditions and other physical causes of disability. But like Herd and her colleagues in the first article, Schoeni, Freedman, and Martin also implicate several more fundamental causes, including greater educational attainment and less poverty among the elderly.

The third article in this issue analyzes the provision of health care for people in the United States who lack health insurance. It is well known that the number of uninsured has been rising in recent years, with almost 47 million people having no health insurance at any time in 2006, according to estimates from the Current Population Survey (DeNavas-Walt, Proctor, and Smith 2007). But it is less well known that some of the ways in which the poor and uninsured can obtain physicians' services have been changing. In “Effects of Changes in Incomes and Practice Circumstances on Physicians' Decisions to Treat Charity and Medicaid Patients,” Peter Cunningham and Jack Hadley analyze some of the reasons why the number of physicians who treat poor patients has been decreasing.

Using data from four rounds of physician surveys in the Community Tracking Study (CTS), the authors found that two factors—physicians' income and ownership of their practice—were associated with physicians' decisions to stop accepting new Medicaid patients or to provide charity care. These factors did not, however, operate in the same way for these two categories of patients. That is, a decline in physicians' income was associated with their decisions to stop accepting Medicaid patients but did not affect their acceptance of charity care patients. Conversely, physicians who owned their practice but then moved to employed positions tended to accept fewer charity care patients but to accept more Medicaid patients. The effect on charity care is important because, as Cunningham and Hadley show, the percentage of physicians who own their own practice has been falling (from 68.9 percent in the 1996/1997 CTS survey to 57.6 percent in the 2004/2005 survey). Unless the number of uninsured persons drops, the declining availability of charity care from private physicians will increase the importance of safety-net organizations that provide medical care.

The final article in this issue returns to a topic that the Quarterly has addressed several times in recent years: ways to increase policymakers' use of research evidence (Jacobson, Butterill, and Goering 2005; Lavis et al. 2002, 2003; Lomas et al. 2003; Mitton et al. 2007). In “Health Research Funding Agencies' Support and Promotion of Knowledge Translation: An International Study,” Jacqueline Tetroe, Ian Graham, and colleagues describe the efforts of thirty-three research funding organizations in nine countries to promote the translation into policy of research that they had supported. The authors studied these organizations, which were mostly government agencies, by interviewing key internal informants about their organizations' efforts to promote the use of research, including requirements placed on investigators. The researchers did not collect data about the policymakers' perspective on the question.

The authors identified eight major types of activities as efforts to “push” information to policymakers, to create a “pull” for information, and to develop “linkage and exchange” relationships between researchers and policymakers. The authors then described the different approaches in detail, including the fact that the respondents used some twenty-nine different terms for what the authors conceptualized as “knowledge translation.” A notable aspect of the researchers' strategy, consistent with the purpose of the project, was sponsoring a workshop for representatives of the Canadian and U.S. agencies in which they could present and obtain feedback about the project's findings. (Funding limitations prevented the participation of agency representatives from the other countries.)

Despite the wide variety of approaches used to promote knowledge translation from researchers to policymakers, the authors concluded that none of the agencies they surveyed could be viewed as “exemplary” in terms of their knowledge translation efforts. No specific activities were required by all of the agencies, and furthermore, the agencies provided little evidence about the effectiveness of different methods. Nevertheless, the article provides an extensive catalog of approaches being used in hopes of enhancing the use of research evidence for policymaking purposes.

I conclude by noting that starting with this issue, The Milbank Quarterly will use a more structured format for abstracts to make them more consistent, informative, and accessible. Guidance for prospective authors about this and other matters continues to be available at