Drawing on Putterman and Weil (2010), we study the impact of deep roots as measured by ancestry by U.S. state, considering the average of how long each state’s ancestors have lived (1) under a centralized state, a variable known as “State History,” and (2) with settled agriculture, a variable known as “Agricultural History.” The other contributions to this literature primarily focus on how State History and Agricultural History of the population affect economic development outcomes across countries. Instead, we look at its effects on economic output per capita across the U.S. States.
Putterman and Weil (2010) created a matrix of contemporary populations of each country based on their population’s ancestral origin in the year 1500, building on earlier work by Bockstette et al. (2002) and Chanda and Putterman (2007). They use a variable, called State History, measuring how long a country has lived under a supra-tribal government, the geographic scope of that government, and whether that government was controlled by locals or by a foreign power. They then discount past State History by reducing the weight on each halfcentury before 1451–1500. For each half-century period, they construct a score, where 0 corresponds to no supra-tribal state, and 50 corresponds to an indigenous supra-tribal state covering most of the present day territory. Intermediate values were defined for foreign or intermittent rule. The State History variable is the discounted sum of the thirty half-century periods, normalized between 0 and 1 (Putterman and Weil 2010: 1640). Their second variable, Agriculture History, measures the number of millennia that have passed since a country transitioned from hunting and gathering to agriculture. Together, we refer to the State History and Agricultural History variables as the “Deep Roots Variables.” Putterman and Weil then combine the matrices of ancestry with the Deep Roots Variables score to show how long each national origin group was governed by a centralized state and how long they had settled agriculture. The Deep Roots Variables score varies dramatically between peoples and locations. According to Spolaore and Wacziarg (2010), under “this approach, the United States has had a relatively short exposure to state centralization in terms of location, but once ancestry-adjusted it features a longer familiarity with state centralization, since the current inhabitants of the United States are mostly descended from Eurasian populations that have had a long history of centralized state institutions.” They find that a country’s Deep Roots Variables scores are positively correlated with GDP per capita today, which suggests that the drivers of economic development and GDP per capita cannot be separated from the deep cultural, historic, and/or genetic roots of human populations. Their findings stand in contrast to those that explain economic development and GDP per capita as the ultimate result of geography, institutions, or other conventional explanations.
Easterly and Levine (2016) find that European ancestry produces a substantial developmental advantage based on data of European settlement. They find that the share of the European population in colonial times has a large and significant impact on income per capita today, even in non-settler colonies governed by extractive economic institutions. Their finding remains large and significant upon controlling for the quality of contemporary institutions. The authors interpret this as consistent with theories favoring human capital as the driver of development. They conclude, “There are many other things that Europeans carried with them besides general education, scientific and technological knowledge, access to international markets, and human capital creating institutions. They also brought ideologies, values, social norms, and so on. It is difficult for us to evaluate which of these were crucial either alone or in combination.”
Our contribution to this literature is to attempt to identify how ancestry affects economic development in the context of a large, industrialized nation. Confining our analysis to differences across U.S. States is proper for two reasons. First, there are wide differences in ethnic and racial heterogeneity across the U.S. States that produce radically different State Deep Roots Variables score. These scores often vary more considerably between U.S. States than between many countries found in the same continent. Second, focusing on the differences between the U.S. states allows us to implicitly control for various other factors that could better explain GDP per capita than deep roots. American states have considerable leeway in managing their own economic institutions and policies within a federal system and have measurably different ancestries. Federalism and different ancestries makes the U.S. States a fertile testing ground for the deep roots hypothesis. To this end, we constructed a matrix of U.S. State ancestries and computed State History and Agricultural History scores for American states, using this ancestry matrix. Next, we compared logged GDP per capita across the U.S. States with the Deep Roots Variables. We additionally compared these scores with the quality of economic institutions that are correlated with levels of logged GDP per capita.
We can only establish very marginal support for Putterman and Weil’s (2010) findings at the state level. Furthermore, there is no statistically significant relationship between the Deep Roots Variables score and the liberalness of a state’s economic institutions. Given the large literature on the importance of liberal economic institutions for economic growth and other outcomes (see, e.g., De Haan et al. 2006; Hall and Lawson 2014), the lack of a relationship between the quality of economic institutions and the Deep Roots Variables eliminates this institutional channel from the deep roots hypothesis.
The structure of this paper is as follows. In Section II, we describe how we apply Putterman and Weil’s methods of measuring the Deep Roots Variables to the U.S. states. Section III compares the Deep Roots Variables to logged GDP per capita and other institutional measures and describes our results. Section IV concludes.
