The proposed economic channel based on child survival and health costs complements other,
socio-cultural explanations of high population growth in Sub-Saharan Africa and other tropical
regions (e.g. Chesnais, 1992). The main advantage compared to explanations built on differences
of preferences is that policy recommendations are straightforward to be drawn from the cost
channel. In short, a health policy accomplishing that all diamonds in Figure 1 are located on
a horizontal line at survival rates of today’s fully developed economies disables the proposed
mechanism for slow or stagnant economic development. Thus, geography’s indirect impact on
development could, at least theoretically, be overcome by improvements of nutrition and by
disease eradication. This slightly optimistic outlook distinguishes the theory from the comple-
menting institutions-based approach where the indirect influence of geography originates from
mortality 200 years ago.
Technological progress has deliberately been modelled as being exogenous because the paper
does not claim to contribute to the new field of “unified growth theories” that explains devel-
opment of countries at the technological frontier (see Galor, 2005). Conversely, the focus on
today’s (non-) developing countries has emphasized that there may exist non-unifiable patterns
of demo-economic development. Depending on geographic location we observe plains and valleys
where there were only hills visible at times of the Western world’s demographic transition and
(unified) growth process.
Nevertheless, interesting future extensions of the model are conceivable. In particular, an
extension towards a two-sector model (as in Hansen and Prescott, 2002) could explain why
decreasing returns to scale are at work at underdeveloped, geographically unfavorable locations
(where demand consists mainly of food) and are insignificant when a demographic transition
has been successfully accomplished and income is high (and demand of industrial products and
services dominates). As a first guess, it can be expected that such an extension would amplify
rather than curtail geography’s indirect influence on economic development.
20