HIV/AIDS, human capital, and economic prospects for Mozambique

As in other countries in the southern Africa region, a human development catastrophe is unfolding in Mozambique. Recently released data estimate HIV prevalence rates amongst the adult population in the year 2000 at around 12% with substantial regional variation. Due to the magnitude of the HIV/AIDS pandemic, it has become a top priority development issue. The goal of this paper is to try to come to grips with the economic dimensions of the pandemic. In this initial assessment, the focus is on implications for macroeconomic prospects. Particular attention is paid to human capital accumulation. Taking demographic projections, human capital accumulation projections, and other inputs, a recursive computable general equilibrium approach is employed to quantify impacts on key macroeconomic variables and identify the major channels through which these impacts occur. Policy options for blunting the major negative impact channels are also considered.

Due to the long time lags between infection and onset of AIDS, the AIDS case projections to 2010 are, barring rapid advance in medical technologies, essentially programmed into the system since nearly all of the people projected to die in this decade, including the latter parts, are already HIV positive. The analysis indicates that these impending AIDS cases and deaths could have large economic impacts. Projecting to 2010, per capita annual GDP growth rates are between 0.3% and 1.0% lower than in a fictional no AIDS scenario. The major sources of this slowdown in growth are (1) reduced productivity growth, (2) reduced population growth and human capital accumulation, and (3) reduced physical capital accumulation. All three of these effects are significant though the productivity effect is the strongest.

Due to a variety of knowledge gaps, a high degree of uncertainty must be associated with these results. However, if AIDS indeed reduces per capita economic growth for extended periods of time as the analysis suggests, then initiatives that effectively combat AIDS will pay handsomely in purely economic terms. Given the nearly decade long time lags between infection and death, policy actions can be divided into two categories: a) reactive policies to face the ramifications of the pandemic in the current decade and b) preventive policies designed to reduce HIV/AIDS prevalence in future decades.

Under reactive policies, education policy was explicitly considered. AIDS poses a threat to educational attainment from both the supply and demand sides. AIDS deaths amongst teachers and administrators will worsen already severe supply side constraints. In addition, widespread orphaning is likely to reduce the demand for education. Both of these effects point to reduced school enrollments and human capital accumulation.

The analysis indicates that school enrollments and human capital accumulation rates are sensitive to changes in the probabilities of remaining in school and transitioning to higher grade levels. The scenario Education, corresponding to a strong policy effort to maintain school enrollment rates, graduation rates, and quality, resulted in a 0.6% increment to GDP growth by 2010 relative to the Base AIDS scenario. This growth increment is due to the enhanced productivity of a more skilled workforce. Furthermore, this increment is likely to persist well into the future due to much larger school enrollments in 2010 (for example, enrollment in EP2 in the Education scenario in 2010 is about twice the level in Base AIDS). Assuming this 0.6% growth increment persists to 2020, net present value calculations justify incremental education expenditures on the order of 5% of GDP per year from 2002 to 2010 in order to obtain the growth increment.

Similarly, if successful preventive policies substantially reduce AIDS deaths in the next decade resulting in a growth benefit of 0.3% per year from 2010 to 2020 (corresponding to the lower end of the estimated average per capita GDP growth impact), net present value calculations also justify large prevention expenditures (2.5% of GDP from 2002 to 2007) even if per capita GDP is the sole criteria for evaluating the preventive policies. Consideration of individual risk of becoming HIV positive and the human costs of AIDS would justify even higher prevention expenditures.

The lesson from these numbers is not that resource allocation on this scale should occur since absorptive capacity is an issue. Rather, the point is that initiatives that successfully prevent the spread of HIV and combat the economic disruptions of AIDS deaths are likely to have a high payoff. Given the extent of the payoff, imaginative initiatives, including costly ones, can be considered.

In the final section, these and other relatively broad policy options are considered. Policies highlighted for further scrutiny include: elimination of school fees (such as fees for textbooks), payments to families in the form of food or cash for enrolling children in school, accelerating the general economic reform agenda, administrative and regulatory simplification, accelerated efforts to extend labor-saving agricultural technologies, and tightly targeted (to teachers for example) use of highly active anti-retroviral therapy (HAART). Widespread use of HAART does not appear to be feasible.

The challenge is in moving beyond general policy ideas (both reactive and preventive) to real policy initiatives. This will require careful thought particularly with respect to issues of implementation.

Arndt, Channing
Published date: 
International Food Policy Research Institute (IFPRI)
Series number: 
PDF file: 
application/pdf icontmdp88.pdf(464.6KB)