Haacker, Marcus, 2002, “The Economic Consequences of HIV/AIDS in Southern Africa,” IMF Working Paper 02/38 (Washington).
World Bank, 2000, “Lesotho: The Development Impact of HIV/AIDS: Selected Issues and Options,” Macroeconomic Technical Working Group (Africa Region) (Washington, World Bank).
Appendix: Model Assumptions; Initial Condition and Parameterization
54. Initial conditions. We take calendar year 2000 to be t = 0, and set k2000 to be a steady state value,18 corresponding to the parameter values for that year (see below). In order to do so, we need the values of the savings rate, s2000, the rate of growth of labor force, n2000, and assumptions about the values of the output share of capital, a, and the annual rate of depreciation of capital, δ. The savings rate is obtained from data reported in the national accounts by the Lesotho Bureau of Statistics. We assume that α = 1/10 and δ = 0.08.
55. Labor force. The growth rate of the labor force is computed using demographic estimates and projections for Lesotho made by the U.S. Census Bureau’s International Populations Center for the period 1985 through 2050. We assume that the labor force is constituted of people between the ages of 15 and 59. We assume a constant labor force participation rate and, consequently, a growth rate of the labor force that is identical to the growth of the population within this age group.
56. HIV/AIDS prevalence and savings rate. UNAIDS estimates that the adult HIV/AIDS prevalence rate at end-2001 is 31 percent. We assume in the model simulations that the prevalence rate among the labor force remains at this level through 2015 and thereafter declines at a constant rate of 1 percent per annum through 2020.19 Savings rates through 2002 reflect data and estimates provided by the Lesotho Bureau of Statistics, and for 2003–08 are based on latest IMF staff projections. From 2009 onward, we assume that the savings rate declines over a five-year period from its 2008 level, reflecting the weaker savings propensities of HIV/AIDS patients and their families. From 2013 onward, we assume that HIV/AIDS patients do not save at all.20
57. Real and Nominal GDP. A benchmark [4.5] percent annual rate of real GDP growth for the medium term in the absence of the epidemic forms the basis for a comparison with an AIDS scenario. In the absence of a sectoral GDP deflator series, nominal GDP for Lesotho is projected forward, using projections of the real GDP growth and the consumer price inflation. Specifically.
Prepared by Jay Surti.
Haacker (2002) in a study of a group of nine Southern African economies estimated that in the medium term, per capita output could be reduced by 4 percent to 10 percent. The estimate for Lesotho was a cumulative reduction of 7 percent over the medium term.
In particular, a comprehensive program of antiretroviral treatment has the potential to significantly lengthen the life span of infected individuals and improve their quality of life. Governments in southern Africa, with the exception of Botswana and South Africa, are not in a position to provide financing for such programs on a wide scale.
See, for example, Haacker (2002) for an application of this method to southern African economies. This paper also contains a comprehensive bibliography of the literature that investigates the savings and labor force channels.
The rate of depreciation of capital, δ, will be assumed constant throughout, as will the share of capital in output, α.
An appendix provides further details on the methodology used to estimate the impact on the savings rate.
We are assuming here that labor is paid its marginal product; i.e., the labor market is competitive.
Through 2008, we adopt the medium-term staff projections of the revenue-to-GDP path, and, for 2009 and beyond, we assume that this ratio grows annually at the average annual percent change between 2003 and 2008.
Based on staff projections through 2008 and for 2009 and beyond, we assume that the annual rate of growth in total expenditure is the average annual percent change between 2003 and 2008.
Hence, the revenue-to-GDP ratio and the level of total government expenditures is assumed to be the same in the AIDS and non-AIDS scenario. Since nominal GDP is lower in the AIDS scenario, the expenditure-to-GDP ratio and the fiscal deficit will be higher in the AIDS scenario than in the non-AIDS scenario.
The long-term numbers are meant to be indicative—as opposed to precise calculation—of the substantially larger impact in the long term over the medium term that may be expected of an epidemic left unchecked.
The steady state value of capital is obtained by holding fixed n and s at their CY 2000
In earlier studies, the World Bank (2002) estimated that prevalence rates among the age group 15–49 would continue to rise through 2010, while Haacker (2002) held labor force prevalence rates constant for his medium-term growth impact analysis.
Our assumptions regarding the savings propensities of HIV/AIDS patients is weaker than Haacker’s (2002), who assumes that they do not save throughout the forecasting horizon. In the case of Lesotho, our weaker assumption may be justified on the grounds that, in the short to medium term, capital formation will largely be driven, as at present, by investments in the textiles industry (financed by entrepreneurs through foreign sources of finance) and in the Lesotho Highlands Water Project (financed primarily through grants from South Africa and Lesotho government financing), both of which are relatively unaffected by the epidemic. However, investment in textiles may reach a saturation point in the medium term, and the water project is expected to wind down by 2008.