References and Literature

The early part of this chapter builds on Acemoglu, Johnson and Robinson (2006), who discuss the distinction between proximate and fundamental causes and the various different 159

approaches to the fundamental causes of economic growth.

The importance of population in generating economies of scale was first articulated by Julian Simon (1990). The model presented in Section 4.2 draws on Simon’s work and the more recent work by Michael Kremer (1993). Kremer (1993) argues for the importance of economies of scale and increasing returns to population based on the acceleration in the growth rate of world population. Another important argument relating population to technological change is proposed by Esther Boserup (1965) and is based on the idea that increases in population creates scarcity, inducing societies to increase their productivity. Other models that build economies of scale to population and discuss the transition of the world economy from little or no growth to one of rapid economic growth include Hansen and Prescott (2001), Galor and Weil (2001), Galor and Moav (2002) and Jones (2004). Some of these papers also try to reconcile the role of population in generating technological progress with the later demographic transition. Galor (2006) provides an excellent summary of this literature and an extensive discussion. McEvedy and Jones (1978) provide a concise history of world population and relatively reliable information going back to 10,000 B.C. Their data indicate that, as claimed in the text, total population in Asia has been consistently greater than in Western Europe over this time period.

The geography hypothesis has many proponents. In addition to Montesquieu, Machiavelli was an early proponent of the importance of climate and geographic characteristics. Marshall (1890), Kamarck (1976), and Myrdal (1986) are among the economists who have most clearly articulated various different versions of the geography hypothesis. It has more recently been popularized by Sachs (2000, 2001), Bloom and Sachs (1998) and Gallup and Sachs (2001). Diamond (1997) offers a more sophisticated version of the geography hypothesis, where the availability of different types of crops and animals, as well as the axes of communication within continents, influence the timing of settled agriculture and thus the possibility of developing complex societies. Diamond’s thesis is therefore based on geographic differences, but also relies on such institutional factors as intervening variables.

Scholars emphasizing the importance of various types of institutions in economic devel­opment include John Locke, Adam Smith, John Stuart Mill, Arthur Lewis, Douglass North and Robert Thomas. The recent economics literature includes many models highlighting the importance of property rights, for example, Skaperdas (1992), Tornell and Velasco (1992), Acemoglu (1995), Grossman and Kim (1995, 1996), Hirsleifer (2001) and Dixit (2004). Other models emphasize the importance of policies within a given institutional framework. Well- known examples of this approach include Perotti (1993), Saint-Paul and Verdier (1993), Alesina and Rodrik (1994), Persson and Tabellini (1994), Ades and Verdier (1996), Krusell and Rios-Rull (1999), and Bourguignon and Verdier (2000). There is a much smaller liter­ature on endogenous institutions and the effect of these institutions on economic outcomes.

Surveys of this work can be found in Acemoglu (2007) and Acemoglu and Robinson (2006). The literature on the effect of economic institutions on economic growth is summarized and discussed in greater detail in Acemoglu, Johnson and Robinson (2006), which also provides an overview of the empirical literature on the topic.

The importance of religion for economic development is most forcefully argued in Max Weber’s work, for example (1930, 1958). Many other scholars since then have picked up on this idea and have argued about the importance of religion. Prominent examples include Huntington (2001) and Landes (2001). Landes, for example, tries to explain the rise of the West based on cultural and religious variables. This evidence is criticized in Acemoglu, Johnson and Robinson (2005). Barro and McCleary (2003) provide evidence of a positive correlation between the prevalence of religious beliefs and economic growth. One has to be careful in interpreting this evidence as showing a causal effect of religion on economic growth, since religious beliefs are endogenous both to economic outcomes and to other fundamental causes of income differences.

The emphasis on the importance of cultural factors or “social capital” goes back to Ban­field (1958), and has recently been popularized by Putnam (1993). The essence of these interpretations appears to be related to the role of culture or social capital in ensuring the se­lection of better equilibrium. Similar ideas are also advanced in Greif (2006). Many scholars, including Veliz (1994), North, Summerhill and Weingast (2000) and Wiarda (2001), empha­size the importance of cultural factors in explaining the economic backwardness of Latin American countries. Knack and Keefer (1997) and Durlauf and Fafchamps (2003) document positive correlations between measures of social capital and various economic outcomes. None of this work establishes a causal effect of social capital because of the potential endogeneity of social capital and culture. A number of recent papers attempt to overcome these difficulties. Notable contributions here include Guiso, Sapienza and Zingales (2004) and Tabellini (2007).

The discussion of the Puritan colony in the Providence Island is based on Newton (1914) and Kupperman (1993).

The literature on the effect of economic institutions and policies on economic growth is vast. Most growth regressions include some controls for institutions or policies and find them to be significant (see, for example, those reported in Barro and Sala-i-Martin, 2004). One of the first papers looking at the cross-country correlation between property rights measures and economic growth is Knack and Keefer (1995). This literature does not establish causal effect either, since simultaneity and endogeneity concerns are not dealt with. Mauro (1998) and Hall and Jones (1999) present the first instrumental-variable estimates on the effect of institutions (or corruption) on long-run economic development.

The evidence reported here, which exploits differences in colonial experience to create an instrumental-variables strategy, is based on Acemoglu, Johnson and Robinson (2001, 2002). The urbanization and population density data used here are from Acemoglu, Johnson and Robinson (2002), which compiled these based on work by Bairoch (1988), Bairoch, Batou

and Chevre (1988), Chandler (1987), Eggimann (1999), McEvedy and Jones (1978). Further details and econometric results are presented in Acemoglu, Johnson and Robinson (2002). The data on mortality rates of potential settlers is from Acemoglu, Johnson and Robinson (2001), who compiled the data based on work by Curtin (1989, 1998) and Gutierrez (1986). That paper also provides a large number of robustness checks, documenting the influence of economic institutions on economic growth and showing that other factors, including reli­gion and geography, have little effect on long-run economic development once the effect of institutions is controlled for.

The details of the Korean experiment and historical references are provided in Acemoglu (2003) and Acemoglu, Johnson and Robinson (2006).

The discussion of distinguishing the effects of different types of institutions draws on Acemoglu and Johnson (2005).

The discussion of the effect of disease on development is based on Weil (2006) and espe­cially on Acemoglu and Johnson (2007), which used the econometric strategy described in the text. Figures 4.10 and 4.11 are from Acemoglu and Johnson (2007). In these figures, initially- poor countries are those that are poorer than Spain in 1940, and include China, Bangladesh, India, Pakistan, Myanmar, Thailand, El Salvador, Honduras, Indonesia, Brazil, Sri Lanka, Malaysia, Nicaragua, Korea, Ecuador, and the Philippines, while initially-rich countries are those that are richer than Argentina in 1940 and include Belgium, Netherlands, Sweden, Den­mark, Canada, Germany, Australia, New Zealand, Switzerland, the United Kingdom and the United States. Young (2004) investigates the effect of the HIV epidemic in South Africa and reaches a conclusion similar to that reported here, though his analysis relies on a calibration of the neoclassical growth model rather than econometric estimation.

4.10.