happened prior to 1980. During the rapid economic growth that occurred after
1980 there was relatively little progress in further reducing child mortality. In
India also growth after 1990 was accompanied by a decline in infant mortality
improvement rates (Drèze and Sen, 2002).
Other evidence of links from growth, such as expenditure on food, is harder to
interpret. Data from a relatively poor, malnourished sample of 240 households in
rural south India in the mid-1970s show that food expenditure increased one-for-
one with income (Behrman and Deolalikar, 1987). This does not, however, imply
that higher incomes will lead to more/better nutrition so improve health. Food
can be purchased for its appearance, status, odour or convenience. Even at low
levels of household income considerable weight is given by households and indi-
viduals to such attributes as taste rather than nutrition in food. The ambiguous
nature of nutrition as a measure of development and its sometimes perverse rela-
tion with economic growth was discussed in Chapter 1.
The social health gradient shows that higher incomes are associated with
higher health status even in developed countries. The two most influential
investigations of the social health gradient in the UK are the two ‘Whitehall
Studies’of British civil servants conducted in 1967 and 1985–88. The first
study found a steep negative relationship between employment grades and
health outcomes, including mortality from various diseases such as heart and
respiratory disease,and cancer. With reasonably well-paid and secure jobs, the
health findings for the civil servants in this sample were not related to unem-
ployment or poverty. The follow-up study (of 10,000 civil servants) in the mid-
1980s found no decline (even a widening in some cases) of the gradient. For
men between 40 and 65 there was a four-fold increase in the risk of mortality
in the lowest relative to the highest employment grade. The gradient remained
despite continued national efforts to equalize health care access for all through
the free National Health Service. The hypothesized explanations for this seem-
ing paradox include the work-related stress, monotony of work and lack of
control over work that may characterize working patterns among lower
employment grades (Smith, 1999:159).
Although there is less empirical work on how growth affects education,
there is some evidence that economic growth increases the expected return
from and so incentives to acquire education (Bils and Klenow, 2000).
From education to economic growth
Education can increase the human capital of the labour force and so boost
labour productivity, increase the innovative capacity of the economy, facilitate
the diffusion and transmission of knowledge and so finally increase GDP. A
theme running through efforts to quantify the impact of education on growth is
the problem of measuring human capital (Mankiw et al.,1992:418–19).
Education investment takes the form of direct spending (by the individual,
family and state) and indirect costs (forgone earnings). Much schooling is
122 Sources of Growth in the Modern World Economy since 1950