Dimensions: 5.5'' x 8.5'' 108 pages AEI Press  (Washington) Publication Date: March 1999 Paperback ISBN: 0844771244 Price: $ 14.95 Add to Cart   Examination Copies

Among wealthy countries, levels of pharmaceutical consumption vary widely. Does greater consumption contribute to better health? If so, can that effect be measured? Answers to those questions would be valuable to policymakers in the health care industry and in government, but the questions are difficult to address. To begin with, analysts must distinguish the health benefits of drug consumption from the benefits of all other forms of health care. The authors of this book undertake that task and analyze data from a sample of twenty-one countries to isolate and measure the health effects of pharmaceutical consumption. Their results show that the use of pharmaceuticals leads to significantly longer lives, especially for those at middle age and beyond.

H. E. Frech III is a professor of economics at the University of California, Santa Barbara, and an adjunct scholar at AEI. Richard D. Miller Jr. is a research analyst with the Center for Naval Analyses in Alexandria, Virginia.

Many international studies of health care are available, especially in the member-countries of the Organization for Economic Cooperation and Development (OECD). Most of those studies have responded to cost-containment problems, and therefore the vast majority have focused on the determinants of health care expenditures.

The research emphasis on estimating the effects of different factors on health care spending is not entirely misplaced, but more effort should be devoted to estimating the determinants of health. Such estimates can help guide policymakers in debates affecting the allocation of resources, both among different types of health care goods and services and between health care and other goods. This book therefore focuses on the productivity of health care, with special attention to disaggregating the effects of pharmaceutical consumption from the effects of all other types of health care consumption.

Pharmaceutical consumption varies considerably from nation to nation, even among the rich countries of Europe, North America, and Oceania. In 1990, for instance, France's per capita consumption of pharmaceuticals was five times that of Denmark, and Italy's was roughly twice that of the United States. A common belief holds that, among rich countries, the marginal return from health care consumption in general, and from pharmaceutical consumption in particular, is negligible. Are those societies that consume more pharmaceuticals acting irrationally, or is there a measurable health return to their higher consumption?

Previous Research

Only a limited number of researchers from a variety of fields, including economics, epidemiology, sociology, and anthropology, have studied the effects of different factors on the production of health. The results of those studies have been mixed, and many of the studies have been flawed. Chapter 2 is a review of that literature.

Certain results of previous studies appear strong and are sensible. First, a few basic public health services, such as a potable water supply and sanitation services, provide the biggest payoffs in decreased mortality for all age groups. Those services, however, are in the domain of civil engineering, not health care. That result has been found by all researchers who have studied underdeveloped countries and who have introduced in their work a sufficient range in those variables of public health infrastructure. Another striking and consistent result is that the expansion of medical care services does not improve mortality rates nearly as much as public health infrastructure development does.

Environmental factors and per capita income have been found to have a much greater effect on mortality than medical care. Higher levels of education are negatively related to mortality. Dietary factors have been found to be important, as richer diets tend to decrease mortality from infectious diseases, although at some point the mortality from degenerative diseases increases as diets become too rich. The same pattern holds true for income. At low income levels, more income is associated with lower mortality rates. At higher income levels (in the most developed countries), income is positively related to mortality rates, at least when education is controlled for. Studies have also found that variations in alcohol and cigarette consumption can explain variation in mortality, while variation in medical care utilization cannot. Finally, several studies have suggested that the political environment may play a role in determining mortality.

While most studies have discovered scant effect on mortality rates from more medical care, the best studies have found small negative effects, that is, increases in longevity. Few studies have dealt with the effects of pharmaceutical consumption on mortality either directly or indirectly. Studies that have dealt with that relationship directly have had serious flaws. One of the better studies--in 1987 by University of Chicago Professor and AEI Adjunct Scholar Sam Peltzman--considered the effects of pharmaceutical regulations on national health indicators and found that mandatory prescription laws are positively related to mortality from poisonings. That relationship may reflect perverse effects of the regulations, or it may reflect merely reverse causation. (The poisonings may have been the reason for such regulations.) Many studies of restricted formularies in the United States have also provided evidence that pharmaceutical consumption has a positive impact on health. The goal of this book was to investigate whether that effect could be found in an international comparison study.

Methods

At the heart of the authors' analysis are mathematical regressions involving multiple variables. Chapter 3 discusses the data and methodological issues concerning measurement of the relevant variables.

The authors use data on twenty-one countries from the 1996 release of the OECD’s health data database. The data include measures of various types of national per capita health care consumption as well as broad measures of health outcomes, such as life expectancies and infant mortality. They also include information on various macroeconomic measures, such as gross domestic product, labor force size, and employment.

As an objective, if crude, measure of the performance of each country’s health care system, the authors use life expectancies as of 1993 at the following ages: birth, age forty, and age sixty. They also use infant mortality as an alternative measure of that performance.

The measure of pharmaceutical consumption used is the 1985 per capita pharmaceutical expenditure for each country, converted to U.S. dollars by using purchasing power parity (PPP) exchange rates for pharmaceuticals from the OECD database. Comparisons across other such measures indicate that the 1985 figures provide a good measure of pharmaceutical consumption. Similarly, the authors construct a measure of health care consumption in 1985 using PPP exchange rates specific to health care. The measures of pharmaceutical and other health care consumption used are the best available for a large number of OECD countries.

The authors control for per capita income, measured as per capita gross domestic product, and other lifestyle factors--tobacco use, alcohol consumption, and richness of diet. Tobacco use is measured as the percentage of the population who smoke, alcohol consumption is measured in liters per capita, and dietary richness is measured as the consumption of animal fat calories per capita.

In the authors' analyses, the explanatory variables are lagged by roughly ten years in the belief that lifestyle factors and medical care consumption will have a cumulative rather than a contemporaneous effect on health. As an example of a cumulative effect, smoking does not kill people immediately but rather does so over a period of years. A full model of this general type would require several lags of each explanatory variable. Both data and sample size limitations lead the authors to include only one lag. First, there are only twenty-one countries in their sample (Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, Turkey, the United Kingdom, and the United States). Second, the PPP conversions for total health care and pharmaceuticals are available only for 1980, 1985, and 1990. Thus, the authors could include additional lagged terms only for 1980 or 1990. Including lags from the 1960s or the 1970s might be worthwhile, but the necessary data do not exist.

Results

The results of the analysis, presented in chapter 4, indicate that per capita pharmaceutical consumption has a significant positive effect on life expectancy, especially at ages forty and sixty, whereas other health care consumption has no measurable effect on life expectancy. The authors also identify factors outside the scope of health care that have a measurable effect on life expectancy, namely, per capita income and richness of diet.

The authors find that pharmaceutical consumption has a positive and significant (both statistically and economically) effect on life expectancy at age forty and at age sixty. It has a small, positive, and statistically insignificant effect on life expectancy at birth.

Furthermore, expressed in terms of dollars per life year saved, the results indicate a large difference between high-use and low-use countries. For men aged forty, for example, estimates range from more than $60,000 per life year saved in France to roughly $3,800 in Turkey. That difference is driven partially by the data and partially by the functional form used. Pharmaceutical consumption appears to have no significant effect on infant mortality, but unfortunately the infant mortality models are not stable.

The authors find that nonpharmaceutical health care consumption has no measurable effect on life expectancy, either at birth, at age forty, or at age sixty. Again, for infant mortality the results are mixed and the models unstable. Gross domestic product is found to have a positive and significant effect on life expectancies at the ages of forty and sixty, although that effect disappears in their European-only sample. The results from the infant mortality regressions are mixed.

Conclusion

This study provides important analysis for policymakers who must allocate resources both among different health care goods and services and between health care and other goods and services. It improves on the existing literature because the authors use better measures of pharmaceutical and other health care consumption and a functional form that allows for diminishing returns. The results are strong and quite stable in the life expectancy models. Even among the wealthy OECD member countries, increased pharmaceutical consumption is surprisingly effective in reducing mortality, especially for those at middle age and beyond.