Education and Trade’s Impact on Macroeconomic Growth: An Example of Non-Linear Regression and Interaction Variables

The role that education and trade has on macroeconomic growth, from a theoretical standpoint, should be positive.  After all, specialization and economies of scale tend to increase with the extent of the market, and education increases the productivity of the labor force which tends to drive real incomes higher.  In order to put the theory to the test, a statistical analysis of these theoretical relationships should be conducted with real data.  Using data from an article published by Levine, Beck and Loayza  and published  in the Journal of Financial Economics such an empirical analysis can be conducted.

The objectives of this post is to answer the following questions:

  1. Will increasing the average schooling in a country increase economic growth?
  2. What impact does increasing the trade share have economic growth?
  3. Does the effect of trade share on economic growth depend on the level of education in a country?

The description of the data, which is available on this website for the textbook, Introductory Economics written by Stock and Watson, are given below:

Click to expand description if image appears blurry.


One of the first things to note is that there appears to be a non-linear relationship between years of school and economic growth, as seen by the image below.  Non-linearity with of per capita GDP and economic growth is also apparent (not pictured), both the education and initial economic endowment will appear in the regression equation transformed monotonically by the natural logarithm.

The regression results with growth as the variable under examination and trade, education, and political stability measures as explanatory variables is picture below.  The amount of trade is positive, but statistically insignificant in generating economic growth.  The average years of schooling on the other hand is statistically significant in explaining growth rates across countries.  A 10% change in average education levels is associated with an increase in the growth rate  of an economy by 18.9%;  if a country grew at 5%, then an increase in the average education level by 10% would have increase the economic growth rate to 6% from 5%.  This might seem like a small difference but given the powerful properties of compounding, this slight increase can make substantial differences over 40 years.  Finally, to answer the third question, there does not appear to be a difference on the economic growth of a country due higher educational attainment that depends on the amount of trade a country does.   All of these results were quantified after controlling for the initial gross domestic product per capita in 1960 and the political stability of  a country measured in with the assassination and revolutions/coups variables.

Interesting results given the theoretical underpinnings of the econometric model which would suggest a positive correlation between traded, education, and economic growth.  The synergistic affects of education interacting with an economy open to international trade were not realized in the data.  Suggestions for future research suggest include looking at panel data or time series data to see if there could be some potential endogeneity or aggregation biased resulting from averaging variables across 35 years.