Memorandum submitted by Simon Dixon,
and Jennifer Roberts
1.1 We recently undertook a preliminary
investigation of the impact of HIV/AIDS on economic growth. The
research used panel data econometrics to estimate a well-established
theoretical model for the growth of GDP per capita. The model
was extended to include the effect of health capital (and consequently
HIV/AIDS) on GDP growth.
1.2 Despite widespread concern with the
effects of the virus, estimates of the impact of HIV/AIDS on economic
growth and performance are sadly lacking. There is even less information
available on the macro-economic costs and benefits of different
government policy responses, the implications for government budgets
and the sectoral and income distribution effects.
1.3 It is essential to study the economic
impact of HIV/AIDS since deteriorations in economic performance
are likely to compound the social and human effects and reduce
the capacity of countries to counter the worst effects of the
1.4 The impact of HIV/AIDS on economic growth
is an empirical question. Theory suggests the epidemic will reduce
productive efficiency, and hence output per worker, reduce the
rate of growth of the labour force and lower the savings rate.
This may cause per capita income and the capital labour ration
to either increase or decrease, because the labour force and savings
rates effects operate in opposite directions in the model.
1.5 Existing empirical findings are seriously
compromised by data limitations and a poor understanding of the
virus in developing countries up to the mid-1990s, which led to
prevalence estimates typically one seventh to one fifth of those
now produced by UNAIDS.
1.6 The main source of data for this research
is the World Bank Global Development Finance and World Development
Indicator databases. Data on HIV prevalence were obtained from
UNAIDS. The results reported here use crudely derived estimates
of the prevalence of HIV between the year of first report and
1997. These will be revised when the embargo on more recent HIV
estimates from UNAIDS expires.
1.7 The data set has a panel structure covering
104 countries (developed and developing) for the time period 1960
to 1998 with data for HIV prevalence, incomes, and physical education
and health capital.
1.8 Health capital is proxied by life expectancy
and life expectancy is in turn affected by the material standard
of living in a country, health care provision and the incidence
of diseasesthe most significant of which is HIV/AIDS. In
this way HIV/AIDS determines economic growth indirectly through
both its impact on population growth and its impact on health
1.9 Exploratory data analysis shows that
those countries with the highest HIV prevalence rates show substantial
downturns in recorded life expectancy (usually in the mid- to
late-1980s). Notable exceptions include South Africa and Swaziland.
1.10 Econometric estimates indicate life
expectancy, the proxy for health capital, has had a substantial
and significant positive impact on the growth of GDP per capita.
1.11 Econometric estimates indicate that
the impact of HIV/AIDS on life expectancy has been significant
1.12 Data problems and some model specification
issues require that these empirical results must be qualified.
1.13 The results reported here provide empirical
support for the arguments that HIV/AIDS is having an impact upon
macro-economic performance in developing countries.
1.14 The growth equations confirm the importance
of life expectancy and physical capital investment to recorded
economic performance, and thereby support the argument that the
economic implications of the epidemic deserve close and careful
1.15 The pronounced impact of HIV/AIDS prevalence
on life expectancies suggest that the epidemic may now be entering
what has been termed stage 6, where the loss of life is starting
to impact appreciably upon social indicators and especially upon
life expectancies. This implies that the economic impact of the
epidemic may accelerate.
1.16 The analyses point up a number of data
issues. There is a need to improve the reliability of education
capital data, and additional information of the prevalence of
life threatening diseases and the provision of health care services
1.17 Further research should be directed
at providing a greater understanding of the mechanisms through
which health, and particularly HIV/AIDS, impacts on national economies.
This research should recognise the differential sectoral impact
of the disease and provide a greater understanding of the mechanisms
by which sustained epidemics impact upon economic performance.
Simon Dixon, Scott McDonald and
13 Lecturer in Health Economics, School of Health
and Related Research, University of Sheffield. Back
Lecturer in Economics, Department of Economics, University of
Lecturer in Health Economics, School of Health and Related Research,
University of Sheffield. Back