Orphaning and HIV/AIDS three analyses from Africa: discussion.

Human capital is at the very core of growth in well-being. Health affects the evolution of the stock of knowledge which in turn affects the evolution of the stock of physical capital and the stock of technological and organizational innovations. All of these factors are known contributors to the wealth of nations, and in recent times, to the global decline in income equality as reported most recently by Sala-i-Martin (2006). All else constant, healthy children have higher school attendance than the less healthy, and thus earning higher wages as adults. Higher adult income is associated with lower fertility rates but higher investments in children's human capital, and so the cycle is repeated. The papers in this session contribute to our understanding of the health-human capital linkage.

The paper by Sharma notes that about 70,000 children become orphans every year in Malawi. The number is over 8% of the total population of the country. She draws upon longitudinal data covering the years 2000-2004 to examine the manner in which orphanhood affects school attendance in rural Malawi. The main result is that the likelihood of dropping out of school is higher for orphans than for nonorphans as the grade level increases and, surprisingly, orphans residing with nonparent caregivers do not have a higher probability of dropping out of school. Orphans under the care of a single caregiver are less likely to attend school. This result may be linked to the household needing to secure an income stream from child labor. The author is left to speculate why orphans are more likely to terminate schooling as grade level increases. Empirical insights into the reason for this result could have important policy implications. The finding of no evidence to support discrimination by nonparent caregivers is also puzzling. Are caregivers altruistic? Are caregivers compensated in some form; perhaps by the expectation of remittances from future earnings of orphaned children in their care, by help with household chores and sibling care that orphaned children might supply, or by remittances from the orphaned children's relatives. Answers to these questions also suggest policy interventions. As always, one can quibble with the econometrics since enrollment status, a choice variable, is likely to be jointly determined with the explanatory variables.

The Arndt et al. paper focuses on whether orphaned children in households headed by a nonbiological adult are discriminated against relative to biological children in the same household. The basic idea is to see if the compression (reduction) in adult good expenditure occurs equally for biological and nonbiological children. Compression suggests an intra-household allocation from adult to child expenditures. They posit a "ratio of equivalent expenditure" to measure the effect of the addition of a child for a given category on relative expenditure. No evidence of discrimination between biological and nonbiological children is found in the nonpoor sample, so focus is placed on the sample for poor households. The key finding is a greater compression of expenditure on adult goods with respect to biological children in eight of the nine instances. The compression concept seems to be based on a static notion of household behavior. If households are viewed as optimizing an inter-temporal utility function subject to an inter-temporal budget constraint, then compression of adult expenditures may not occur. Instead, savings may adjust so that a smaller share is allocated to physical assets and a larger share allocated to investments in children, leaving the total level of savings and expenditure on adult goods unchanged. It is possible for the entire share of the reallocated savings to be invested in biological children, with no allocation to the nonbiological and yet no compression in adult expenditures. The observation of compression in relative expenditures could be a "timing phenomenon," or confounded by the allocation of savings between biological and non-biological children. The authors found no evidence of discrimination in the full sample, nor in the non-poor sample. This result is a potentially important finding if it can be verified. It suggests that incomplete markets and market failures are likely a greater burden on the poor than on wealthier households. In the context of this study, this burden falls disproportionately on orphans. Human capital, in contrast to capital embodied in other animals, has always been an asset that is difficult to collateralize. Government interventions targeted to collateralize the human capital of orphans in low income households could potentially earn very high social returns, and even have positive spillovers to other siblings.

The Beegle et al. paper seems carefully crafted and draws upon an exceptional Tanzanian two-period (early 1990s, and 2004) data set. This data set has the potential to make a major contribution to our understanding of the effects of orphanhood on human capital formation. The 10-15 year time span between the two surveys of the same individuals allows them to assess the effects of orphanhood shocks when individuals were young and fall into the 11-18 years of age in 2004. They can also assess the effects of shocks when individuals were in the 10-15 age category in the early 1990s and now fall into the 19-28 age group in 2004. The finding of orphanhood shocks affecting the 19-28 age group category can be inferred as having permanent effects. Some of their results pose questions left to be answered. For example, why do older children fare worse when orphaned from a richer baseline household? While the significance of some of their estimated coefficients might be capturing the effects of unobservable covariates, the overall results are stimulating and provide important insights into orphanhood and human capital formation.

I conclude with two summary remarks. The first is: where is the economics in these papers? Could the papers have been written by statisticians with no training in economics? Granted, authors hint at "important policy implications" and space constrains the development of a conceptual framework. Nevertheless, the social value of this work depends on key next steps that involve how these findings link to the overall behavior of households, the evolution of human capital, and how the presence of markets and organizations can ameliorate the human capital effects of orphanhood. The second remark involves the "bigger picture." What is the cost to an economy over time of orphanhood, or the trade-off between allocating resources to mitigate the negative effects of orphanhood directly and investing in HIV/AIDS mitigation? Alemu, Roe, and Smith (2002) find that HIV/AIDS can have a large negative impact on total factor productivity (TFP). They find that disease prevalence causes a decline in TFP of about 23% for Lesotho, and 15% for South Africa. Incorporating South Africa's epidemiological model of HIV/AIDS dynamics into a neoclassical growth model with endogenous savings, Roe and Smith (2005) find that over the 2004-2020 period, the presence of the disease will only decrease per capita GDP by less than 0.2 percentage points. However, the effect on the level of output is staggering, causing GDP to be about half the value it would be without the disease after about three generations. Efforts need to be made that incorporate the kind of microeconomic results of the papers presented here into the economy-wide models in order to capture the overall effects of the disease on well-being, and to help design and target economic policies to mitigate its effects on the broader economy.

References

Alemu, Z., T. Roe, and R. Smith. 2002. "The Impact of HIV on Total Factor Productivity." Paper presented at the Minnesota Economic Development Conference, 28-29 April.

Roe, T., and R. Smith. 2005. "Incorporating Epidemiological Projections of Morbidity and Mortality into an Open Economy Growth Model: AIDS in South Africa." Invited Paper, African Development and Poverty Reduction: The Macro-Micro Linkages, Cape Town, South Africa, 13-16 October.

Sala-i-Martin, X. 2006. "The World Distribution of Income: Falling Poverty and Convergence, Period." Quarterly Journal of Economics 121(May):351-97.

Terry Roe is a professor in the Department of Applied Economics, University of Minnesota.

This article was presented in a principal paper session at the AAEA annual meeting (Long Beach, CA, July 2006). The articles in these sessions are not subjected to the journal's standard refereeing process.