Keywords

JEL Classification

Population health and a high level of income go hand in hand. Higher incomes promote better health through improved nutrition, better access to safe water and sanitation, and increased ability to purchase more and better quality health care. There is also, however, an effect of health on income. This can work through several mechanisms (Bloom and Canning 2000). The first is the role of health in labour productivity. Healthy workers lose less time from work due to ill health and are more productive when working. The second is the effect of health on education. Childhood health can have a direct effect on cognitive development and the ability to learn. In addition, because adult mortality and morbidity (sickness) can lower the prospective returns to investments in schooling, improving adult health can raise the incentives to invest in education. The third is the effect of health on savings. A longer prospective lifespan can increase the incentive to save for retirement, generating higher levels of saving and wealth, and a healthy workforce can increase the incentives for business investment. We examine the evidence for these mechanisms and find that there are potentially large effects of health on economic outcomes at both individual and macroeconomic levels.

Improved population health has a large impact on population numbers and age structure, and we examine the economic implications of this induced demographic change. The global population explosion of the nineteenth and twentieth centuries was caused not by a rise in fertility but by a fall in mortality. Lower mortality and improved survival rates increased population numbers, but also led to significant increases in the number of young people since the largest improvements in mortality are initially in infant mortality rates. In the long run, reductions in infant mortality lead to a fall in desired fertility, creating a one-time baby-boom cohort. As this large cohort ages, the resultant changes in population age structure can have significant economic implications.

The issue of population health and economic outcomes is particularly acute in sub-Saharan Africa. This region has a high burden of tropical and other infectious disease, such as malaria, tuberculosis, and intestinal worms, and it also suffers from the HIV/AIDS pandemic. We examine the impact of this disease burden on the prospects for economic development in sub-Saharan Africa.

Although we focus on the economic implications of population health, there is clearly two-way causality as health is partly a consequence of income levels. Preston (1975) demonstrated a positive correlation between national income levels and life expectancy. One reason for this link is that higher income levels allow greater access to inputs that improve health, such as food, clean water and sanitation, education, and medical care. Fogel (2004) emphasizes the role of access to food while Deaton (2006) puts more weight on public health measures such as clean water and sanitation (see Cutler and Miller 2005). Cutler and McClellan (2001) examine the increasing contribution of medical care to health outcomes. Pritchett and Summers (1996) use the relationship between income levels and health to argue for an emphasis on economic growth in poor countries as a method of increasing population health. However, the findings of Easterly (1999) weaken this argument. Easterly finds that, although income levels and population health are closely related, the effect of changes in income on population health over reasonable time spans appears to be quite weak. By contrast, relatively inexpensive public health interventions and policies can have remarkable impacts on population health even in very poor countries. In practice, the major force behind health improvements has been improvements in health technologies and public health measures that prevent the spread of infectious disease, and not higher incomes (Cutler et al. 2006).

We examine the role of health as an instrument to generate economic wellbeing. However, any reasonable view of the contribution of health to human welfare would also include the direct welfare benefits of a long lifespan and good health. Estimates of the monetary value of life (as measured by the willingness to pay to avoid a small risk of death) are often very large (Viscusi and Aldy 2003). We can use these estimates of the value of life to compare the welfare improvements that have come about due to improvements in population health and the improvements due to economic growth and higher incomes. Such comparisons suggest that in many countries the value of health gains has been comparable to, or has even surpassed, the value of income gains (Nordhaus 2003; Becker et al. 2005).

Health as Human Capital

The idea of health as a form of human capital has a long history (for example, see Mushkin 1962). Grossman (1972) develops a model in which illness prevents work so that the cost of ill health is lost labour time. However, there may also be an effect of ill health on worker productivity in employment. A major difficulty in measuring the economic effect of health is the two-way causality between wealth and health (Smith 1999). Another difficulty is the lack of consensus on what is meant by health. Different studies use different health measures: self-assessments of health, biomarkers, medical records, limitations on physical functioning, and anthropometric measurements have all been used as health indicators. Each of these approaches may fail to provide a complete picture of an individual’s health status, giving rise to a problem of measurement error. In addition, it is necessary to separate out the effect of investments in health from the effect of natural or genetic variation in health (Schultz 2005).

One solution to these problems in measuring the effect of health on worker productivity is to establish the causal paths in panel data through the use of timing of health shocks and income or wealth responses (for example, Adams et al. 2003). Case et al. (2005), controlling for parental influences and education, find that childhood health has a significant impact on adult health and earnings. Yet another approach to establishing causality is to use instrumental variables. For example, Schultz (2002) instruments adult height with childhood health and nutrition to argue that each centimeter gain in height due to improved inputs as a child in Ghana and Brazil leads to a wage increase of between 8 and 10% (Strauss and Thomas 1998, provide a survey of studies in this area).

Thomas and Frankenberg (2002) caution against drawing inferences from observational studies and instead advocate an experimental approach. A randomized experiment using iron supplementation to reduce iron deficiency anemia led to sizeable effects on worker productivity in Indonesia (Basta et al. 1979). Quasi-experiments can be used where it is possible to treat changes to health as if such changes were randomly generated. Bleakley (2003) considers the effects of the eradication of hookworm and malaria in the United States in the 1910s and 1920s. These diseases were pandemic in many counties of the American South prior to eradication. Bleakley, controlling for normal wage gains in areas that were not infected, shows that children not exposed to these diseases due to their eradication had improved incomes as adults relative to those born before eradication.

This body of research on health and human capital generally supports the idea that health affects worker productivity. However, it lacks a good appreciation of which types of health intervention are most important and what rate of return can be achieved by investing in health as a form of human capital. In many developing countries, relatively inexpensive activities designed to prevent the spread of infectious disease (for example, vaccination) can increase population health at low cost, suggesting that even modest income gains from health will generate very high rates of return. By comparison, treating chronic non-infectious disease in developed countries is often costly. There is evidence that susceptibility to chronic disease in later life is determined by health and nutrition as a fetus and in infancy (Barker 1992; Behrman and Rosenzweig 2004), suggesting that early health investments are crucial for adult productivity.

Health and Education

Education is widely agreed to affect economic outcomes, and health affects education through two mechanisms. The first is the effect of better child health on school attendance, cognitive ability, and learning. Bleakley (2003) finds that deworming of children in the American South had an effect on their educational achievements while in school. Miguel and Kremer (2004) find that deworming of children in Kenya increased school attendance.

The second mechanism is the effect of lower mortality and a longer prospective lifespan on increasing incentives to invest in human capital. This effect occurs for the individual for whom the benefits of education are now greater (Kalemli-Ozcan et al. 2000). In addition, lower infant mortality may encourage parents to invest more resources in fewer children, leading to low fertility but high levels of human capital investment in each child (Kalemli-Ozcan 2002). Evidence for this effect is limited, though Bils and Klenow (2000) do find an effect of life expectancy on investments in education at the national level.

Health and Saving

Poor health affects both the ability to save and the impetus to save. Sickness can have a large effect on out-of-pocket medical expenses, which can reduce current and accumulated household savings. This occurs in developed countries (Smith 1999) but is of particular concern in developing countries where families may be thrown into poverty if productive assets such as land or animals must be sold to pay for medical expenses.

Because poor health tends to be associated with a short lifespan, increasing population health and expected longevity will have an effect on the planning horizon and will influence life-cycle behaviour. With a fixed retirement age, a longer lifespan elicits greater savings for retirement. Blanchard (1985) considers the theoretical effect of a longer lifespan in a macroeconomic model. Hurd et al. (1998) find that increased expectation of longevity leads to greater wealth-holding at the household level in the United States. Bloom et al. (2003) find an effect of life expectancy on national savings, using cross-country data. Lee et al. (2000) argue that rising life expectancy can account for the boom in savings in Taiwan since the 1960s. But the effect of a longer lifespan need not be increased saving for retirement; people could instead choose to work longer. The behavioural response to longer lifespans depends on social security arrangements and retirement incentives (Bloom et al. 2007).

In a life-cycle model with a stable age structure and no population growth or economic growth, the dissaving of the old will exactly match the saving of the young at any level of life expectancy. This suggests that the aggregate effect of longer lifespans on savings is temporary and occurs when life expectancy rises. In the long run, the high savings rates of the working age population will be off set by the dissaving of a large cohort of elderly.

An effect on saving may lead to higher investment if capital markets are not perfectly open. In addition, a healthy population and workforce may increase productivity and encourage foreign direct investment (Alsan et al. 2006).

Health and Demography

Improvements in health and decreases in mortality rates can catalyse a transition from high to low rates of fertility and mortality – the ‘demographic transition’ (Lee 2003). Population growth is the difference between birth and death rates (ignoring migration) and the global population explosion in the twentieth century is attributable to improvements in health and falling death rates. In developing countries, health advances tend to lower infant and child mortality rates, leading initially to a surge in the number of children. Reduced infant mortality, increased numbers of surviving children, and rising wages for women can lower desired fertility (see Schultz 1997) leading to smaller cohorts of children in future generations. Better access to family planning can also help couples achieve match more closely their fertility desires and realizations. This process creates a ‘baby boom’ generation that is larger than both preceding and succeeding cohorts. Subsequent health improvements tend primarily to affect the elderly, reducing old-age mortality and lengthening the lifespan.

In many theoretical models a population explosion reduces income per capita by putting pressure on scarce resources and by diluting the capital–labour ratio. In these models population declines spur economic growth in per capita terms. For example, the very high death rates, and decline in population, due to the Black Death in fourteenth century Europe appear to have caused a shortage of labour, leading to a rise in wages and the breakdown of the feudal labour system (Herlihy 1997). However, in modern populations there appears to be little connection between overall population growth and economic growth; indeed the twentieth century saw both a population explosion and substantial rises in income levels.

Although it is difficult to find significant effects of overall population growth on economic growth, it is possible to consider the components of population growth separately. High birth and low death rates both generate population growth, but seem to have quite different effects on economic growth (Bloom and Freeman 1988; Kelley and Schmidt 1995). This may be because, while both forces increase population numbers, they affect the age structure quite differently. The effect of changing age structure due to a baby boom has large effects as the baby boomers enter the workforce and then as they eventually retire. While the baby boomers are of working age, economic growth may be spurred by a ‘demographic dividend’ if the baby boom generation can be productively employed. Bloom et al. (2004) find that the demographic dividend increases the potential labour supply but its effect on economic growth depends on the policy environment.

There is a worry that health improvements and population aging will lead to high dependency rates and a slowdown in economic growth. In addition to longer lifespans, however, we are seeing a compression of morbidity; the period of sickness towards the end of life is falling as a proportion of overall lifespan (Fries 1980, 2003). The idea that old-age dependency starts at 65 is essentially a result of social security retirement arrangements (Gruber and Wise 1998) and healthy aging means that physical dependency now often occurs at much later ages.

Health and Economic Growth

In growth models, population health is usually taken to be life expectancy, or some other mortality measure, as opposed to the morbidity measured used at the individual level. This disjunction can be bridged by assuming a one-to-one relationship between mortality and morbidity rates in a population; however it is not clear that such a relationship holds, making comparison of the macroeconomic relationship and microeconomic relationships difficult. In addition, calculating life expectancy requires age-specific mortality rates that are unavailable for many developing countries and published life-expectancy figures from the World Bank and United Nations are often constructed from quite incomplete raw data (Bos et al. 1992). There is a need to improve our measures of population health and to expand them to measures that correspond to morbidity and not just mortality.

The effect of health on individual productivity implies a relationship between population health and aggregate output. Shastry and Weil (2003) calibrate a production function model of aggregate output using microeconomic estimates of the return to health. They find that cross-country gaps in income levels can be explained in part by differential levels of physical capital, education, and health, with these three factors being roughly equal in terms of their contribution to differences in income levels. (A little over half of cross-country income gaps are explained by these factors; the remainder of the gap is ascribed to differences in total factor productivity.)

Another approach estimates the effect of population health on economic growth. Estimating the effect of the current level of population health on current income levels is subject to the problem of reverse causality; income also affects health. One way around this problem is to look at the effect of population health on subsequent economic growth, arguing that the timing can determine the direction of causality. This requires the absence of reverse causality through an expectation effect (so that current health is not caused by expected future economic growth).

Growth regressions show that the initial levels of population health are a significant predictor of future economic growth (Bloom et al. 2004, provide a survey of this literature). Sala-i-Martin et al. (2004) find that the predictive power of health (as measured by life expectancy and malaria prevalence) is robust to the specification of the growth regression. Bhargava et al. (2001) argue that the effect of health on economic growth is larger in developing countries than in developed countries.

While population health measures are highly predictive of future economic growth, there is a debate about how to interpret the link. The health effect could be interpreted as the macroeconomic counterpart of the worker productivity effect found in individuals. However, Acemoglu et al. (2003) argue that health differences are not large enough to account for much of the cross-country difference in incomes, and that the variations in political, economic and social institutions are more central factors. They argue that health does not have a direct effect on growth, but serves in growth regressions as a proxy for the pattern of European settlement, which was more successful in countries with a low burden of infectious disease.

Even if a causal interpretation of the effect of health on individual productivity and economic growth is accepted, the argument for using health as an input depends on there being low-cost health interventions that can increase population health without first having a high income level. There are, however, a large number of such interventions that can be implanted (Commission on Macroeconomics and Health 2001).

Tropical Disease and HIV/AIDS

Sub-Saharan Africa suffers from poor health due to the widespread presence of tropical disease. Malaria and tuberculosis cause high illness and death rates, while parasitic diseases such as schistosomiasis and intestinal worms can cause anemia and reduced energy levels and productivity. In addition to these tropical diseases, the high prevalence of HIV/AIDS is causing life expectancy to decline dramatically in many countries in the region. Poor health status is one cause of sub-Saharan Africa’s economic stagnation (Bloom and Sachs 1998). Malaria appears to have an effect on economic growth over and above that created through higher mortality, suggesting that its effects on productivity with a given mortality burden are greater than other diseases (Gallup and Sachs 2001).

Although HIV/AIDS has increased mortality rates dramatically, its impact on income per capita is unclear. HIV/AIDS is associated with high mortality but the period of sickness before death is relatively short. This mutes the worker productivity effects of the disease. Bloom and Mahal (1997) find that HIV/AIDS does not seem to lower the growth rate of income per capita; lower output is matched by lower population numbers due to high death rates. Young (2005) goes further and argues that AIDS mortality reduces fertility significantly, and that this will lower population pressure and increase the income per capita of the survivors of the pandemic in South Africa.

Many authors, however, argue that AIDS mortality has significant indirect effects that will reduce economic growth in the long term. Deaths from HIV/AIDS are concentrated among young adult men and women, leading to a higher dependency ratio. Bell et al. (2004) argue that the creation of a generation of AIDS orphans may lead to lack of care and education for children and to low productivity in the future. This effect may be compounded by fatalism induced by high AIDS mortality and shortened expected lifespan, which reduce the return to education. The high level of stigma associated with HIV/AIDS can reduce trust in the community, while high mortality and the strains imposed by extreme ill health before death can weaken families, community groups, firms, and government agencies, with long-term consequences for social capital (Haacker 2004).

It is important to remember that income per capita is not a complete measure of welfare. Resources devoted to preventing and treating HIV/AIDS are part of measured income but reduce consumption of other goods, reducing welfare even as measured GDP per capita may remain steady (Canning 2006). A more comprehensive welfare measure that included the welfare gain derived from a long lifespan, as well as annual income, would show a large welfare reduction due to HIV/AIDS (Crafts and Haacker 2004). The main welfare effect of HIV/AIDS is the sickness and death of its victims and the impact of these on the victims’ families; the effect on the average income level of the survivors is decidedly secondary.

See Also