On March 10, 2012 the New York Times published an article by columnist Thomas L. Friedman entitled “Pass the Books. Hold the Oil.” The article highlighted a study by the Organization for Economic Cooperation and Development (OECD) which mapped a correlation between performance on the Program for International Student Assessment (PISA)- which every two years tests math, science and reading comprehension skills of 15-year-olds in 65 countries- and the total earnings on natural resources as a percentage of GDP for each participating country. The results revealed a significant and negative relationship between the money countries receive from natural resources and their PISA scores in mathematics.
This article inspired me and a collegue of mine to do a little research that ultimetely turned in to a conference paper. In this post, I will present our conceptual framework. In later posts, I will present our methodology, findings, and analysis.
Our Conceptual Framework:
The relationship between natural resource dependency and education is revealed within a broader body of literature examining the relationship between natural resource dependency and economic growth. This body of literature has shown that economies with a high ratio of natural resource exports to GDP have tended to grow slowly (Sachs, 1997). Explanations for this inverse relationship, to date, have coalesced into a model based upon five main channels of transmission: the Dutch Disease phenomena; rent seeking and social capital; education and human capital; saving, investing, and physical capital; money, inflation, and financial capital (Gylfason, 2004). For the purposes of this paper, we will focus almost exclusively on the relationship between natural resource dependency and education and human capital.
The Dutch Disease, a concept that explains the relationship between the increase in exploitation of natural resources and a decline in the manufacturing sector, has existed within macroeconomic theory since the 1960’s (Neary, 1986). To better conceptualize the Dutch Disease phenomena, contemporary economic historians have stressed the lack of positive externalities associated with natural resource sectors as a possible explanation for the inverse relationship between natural resource dependency and economic growth. Hirschman (1958), Seers (1964), and Baldwin (1966) posit that manufacturing, as opposed to natural-resource production, leads to a more complex division of labor and hence to a higher standard of living (Sachs, 1997). In Matsuyama (1992), manufacturing is characterized by learning-by-doing that is external to the firm. Hence the social return to manufacturing employment exceeds the private return and leads to human capital accumulation in the economy. Matsuyama’s model also proposes a zero-sum assumption, arguing that any demand for employment in the non-manufacturing sector reduces employment that otherwise would occur in the learning-by-doing manufacturing sector, resulting in adverse effects on economic growth. The presence of natural resource dependency thus leads to “Dutch Disease.” Dutch Disease models demonstrate that the existence of large natural resource sectors will affect the distribution of employment throughout the economy (Sachs, 1997).
Taking this argument a step further, Sachs and Warner state the negative effect of large resource endowments on growth need not be dependent on the presence of production externalities in manufacturing, but instead could result from increasing returns to scale in education or job training. In this scenario, a young person incurs the costs of education only if he or she expects to be employed in the manufacturing sector. Thus, the sectoral composition of the economy serves as a signal to citizens. This effect is further compounded when we assume that the education production function is a multiple, greater than one, of the skill level of the teacher. Consequently, a natural resource dependent economy can become stagnant as each generation chooses to forgo education. In contrast, a resource-poor economy, which employs a large percentage of its population in manufacturing, will produce incentives to invest in education. A virtuous circle of endogenous growth results from this as the education process produces not only more skilled workers but more skilled teachers in the next generation (Sachs, 1997). According to Gylfason (2004), this linkage reinforces the case for investment in education and training as an engine of growth as more and better education tends to shift comparative advantage away from primary production towards manufacturing and services, and thus to accelerate learning by doing and growth.
Gelb (1988) stressed that governments typically earned most of the rents from natural resource exploitation. The rent seeking behavior associated with natural resource exploitation can create disincentives for investing in areas that may diminish natural resource exploitation. Matsuyama’s model predicts a zero-sum game between employment in manufacturing and employment in natural resources sectors. Thus, governments that receive a disproportionate share of their revenue from natural resource rents will be less inclined to invest in an education system that will pull workers from the natural resource sector and push them in to manufacturing.
As outlined above, contemporary models identify education and human capital as one of the five main channels of transmission in the inverse relationship between natural resource dependency and economic growth. From these models we adopt the empirically derived notion that natural resource dependency can have negative effects upon education and human capital accumulation. However, under certain circumstances, a high level of human capital can offset the negative effects of natural resource dependency. Countries whose levels of human capital more than offset the expected negative effect of natural resource dependence on growth were found in a cross-country analysis (Bravo-Ortega, 2005). To broaden our framework, we reviewed labor economics literature in search of a theoretical framework surrounding the growth effects of education in and of itself. In this literature we found parallels to Sachs and Warner’s virtuous circle concept. Uzawa (1965) and Lucas (1988) said that, in the long term, sustained growth is only possible if human capital can grow without bound. Bils and Klenow (2000) suggested that this could occur if the quality of education increases over time. This compounding return will make a difference in productivity in later employment (Temple, 2001). In regards to the hypothesis that human capital accumulation can have a compounding effect upon economic growth, there seems to be relative agreement among the various bodies of literature. Bravo-Ortega’s contribution comes from the finding of statistical evidence of a positive relationship between human capital and economic growth after controlling for natural resource dependence.
In summary, our framework was extrapolated from a larger body of research analyzing the relationship between natural resource dependency and economic growth. Contemporary models predict five main channels of transmission. For the purposes of formulating a conceptual framework for our research, we decided to focus exclusively upon the theoretical relationships between natural resources and economic sectoral composition, the ways in which that composition signals workers to either forgo or invest in education, the compounding macroeconomic results of the decisions to invest in or forgo education, the disincentives for government investment in education produced by rent seeking behavior, and the circumstances under which human capital accumulation can offset the negative effects of natural resource dependency on economic growth. It should also be noted that we have limited our definition of “natural resources” to those resources used in the production of energy, more specifically those derived from fossil fuels- oil, coal, and natural gas.
This completes the discussion regarding the conceptual framework from which we began our inquiry. In following posts, I will discuss the methodology we used to select and collect various state-level education and natural resource dependency variables for all 50 states, as well as the statistical methods we used to examine the relationships between all variables.