source: Hay, Guerra, Gething, Patil, Tatem, Noor, et al. (2009).
Another important geographic factor is disease burden. Countries in the “malaria belt” in sub-Saharan Africa continue to be underdeveloped (see Figure 2.2). Malaria has probably killed more human beings over the course of human history than any other disease. But the burden of malaria is also economic. All else being equal, countries where a high proportion of the population are infected with malaria had growth rates that were around 1.3% lower than other countries (Sachs and Malaney, 2002). They also have higher infant mortality and lower investment in physical and human capital.
Other diseases are also endemic in sub-Saharan Africa. One particularly potent disease is sleeping sickness, which the tsetse fly transmits via parasites. The effects of the tsetse fly are not limited to humans. The same parasite causes nagnana, which is deadly to livestock. Livestock played a crucial role in agricultural development in other parts of the world. People in regions affected by the tsetse fly were historically much less likely to domesticate livestock. Alsan (2015) documents how in these parts of sub-Saharan Africa, intensive agriculture was less likely to develop, the plow was less likely to be employed, and large domesticated animals were rarer. These patterns of economic underdevelopment also had political consequences, as they made it less likely for centralized states to develop.
These examples point to the attractions and limitations of arguments based on geography. On the one hand, geography-based arguments are simple and geography has the advantage of being largely exogenous. This means that it is not affected by other variables of interest. As such, we don’t have to worry about geography being the result of other factors that are also important for economic growth, such as culture or institutions. Hence, geographic explanations can potentially provide a straightforward explanation of economic growth and poverty.
The big problem with geographic explanations, however, is that geography is largely unchanging. Geography is not completely static, of course. For instance, fisheries can be over-fished, and this encourages societies to adapt accordingly (Ostrom, 1990; Dalgaard, Knudsen, and Selaya, 2020). And natural resources are generally not just there for the taking – finding them requires exploration and the capacity to extract (David and Wright, 1997). But geography is much less malleable than other societal features such as demography, institutions, or culture. This can be a problem for explanations linking geography to long-run economic growth, because many of the differences in incomes that we observe across the world today have changed dramatically over time. In 1750, the richest country in the world (per capita) was probably the Dutch Republic. Yet, it was only at most four times richer than the poorest countries in the world. Today the richest countries in the world are between one hundred and two hundred times richer in terms of measured GDP per capita. More perplexing still, geography has a difficult time explaining economic reversals. As it is mostly fixed, it cannot easily explain why the Middle East was much more developed in 1000 than Western Europe yet by 1800 Western Europe was far ahead of the Middle East.
Guns, Germs, and Steel
In his Pulitzer Prize-winning book Guns, Germs, and Steel, Diamond (1997) asked why it was that Europeans were able to conquer the New World so easily. How was it that they had not only superior weapons and technology but also more deadly germs than the native inhabitants of the Americas? His answer was geography. To explain why Eurasian societies were more technologically advanced, he drew on the work of Crosby (1986), who argued that the relative height and length of the Eurasian, African, and American continents had a deep impact on long-run development. Vertically aligned continents contain numerous microclimates, which limit the spread of crops, domesticated animals, and people. After all, crops suitable for rain forests are unlikely to grow in the savannah or mountains. So, the domesticated plants and animals of Mesoamerica did not spread to Peru or the Amazon basin. Perhaps more importantly, technology and knowledge spread more easily horizontally than vertically, since climatic characteristics vary less. The “verticality” of the Americas and Africa south of the Sahara (see Figure 2.3) may have therefore imposed huge hurdles to long-run economic growth.
Figure 2.3 Verticality and horizontality of the continents
Reproduced using map in Diamond (1997). Gall-Peters map projection.
This, according to Diamond, was why the Near East was the birthplace of human civilization. It was in the Near East that wheat, barley, and the pea were first cultivated. The rich fertile plains between the Tigris and Euphrates rivers were surrounded on several sides by so-called “hilly flanks.” Goats and sheep were first domesticated 11,000 years ago in the hills of northern Syria and northern Mesopotamia. Pigs and cattle followed soon afterwards and spread throughout the Fertile Crescent. In total, thirty-three of the world’s fifty-six heaviest wild grasses originated in either Europe, the Near East, or North Africa. The majority of these crops were first domesticated in the Fertile Crescent. Moderate climatic differences between Europe and the Middle East ensured that agricultural techniques could spread between the two.
Of the fourteen species of large domesticable animals, only one, and perhaps the least useful – the llama – originated in the New World. Nine of the world’s fourteen large domesticable herbivores originated in Eurasia. This left inhabitants of the New World at a decided economic disadvantage. Livestock are a tremendous source of both animal power and protein. But domesticated animals are also a source of epidemic disease. The absence of large domesticable animals left New Worlders particularly vulnerable to new diseases. Because interactions between humans and animals are a major source of epidemic disease, human populations which have a long history of exposure to these diseases are more likely to acquire immunity. This is why contact with Europeans was so deadly for Native Americans. Recent estimates suggest there was a 95–98% drop in the Native American population between its pre-1492 peak and 1900 (Mann, 2005).
Figure 2.4 Technology adoption levels in 1500 (% of frontier technologies adopted)
Data source: Pavlik and Young (2019).
Diamond’s arguments about the diffusion of technologies across regions have received empirical support. Pavlik and Young (2019) find that technologies moved more easily between east–west neighbors than between north–south neighbors. This is plainly seen in Figure 2.4, which maps how close to the technological frontier different parts of the world were in 1500. On the eve of the European discovery of the Americas, the technology of the Eurasian east–west axis was clearly more advanced than that of the American north–south axis or that of the sub-Saharan Africa north–south axis. These differences would only become exacerbated with the onset of colonization (see Chapter 6).
Mountains, Coasts, and Climate
Markets are unlikely to exist where it is difficult to transport goods and people. Geography is thus an important determinant of market size today (Redding and Venables, 2004). It mattered even more in the pre-industrial period. Prior to the spread of the railroad in the mid-19th century, water transport was by far the most cost-effective way to transport goods from one place to another. It was often twenty times cheaper to transport a good by sea than by land. So, it is reasonable to suspect that access to navigable waterways was historically an important component of economic growth.
Venice
