**NSA Lecture in Statistical Methods **

**Case Study on the Impact of Defense Expenditures in Developing and Emerging Economies**

**Overview**

It is common in economics classes to discuss the guns vs. butter problem. Specifically, in a resource constrained economy the more of either results in reduced amounts produced in the other. Does this mean that countries that increase their defense expenditures do so at the sacrifice of their production of other goods and thus lower their standards of living?

An initial study at Colombia University by Emile Benoit using regression analysis on a large number of developing and emerging economies found little or no stastical verification of the guns vs butter trade-off -- i.e., defense expenditures were insignificant in affecting economic growth rates.

Research at the Naval Postgraduate School extended Benoit's original work and found that sub-groupings of countries produced a significantly different result. The methods used in this analysis involved multiple regression analysis, factor analysis and discriminate analysis. This analysis initially split countries into two broad groups -- those constrained by resource shortages and those relatively unconstrained. For the constrained countries, defense expenditures remained insignificant in affecting growth after other control variables were taken into account. For the unconstrained countries, defense expenditures had a positive and statistically significant impact on economic growth

The readings below start with a brief introduction to regression analysis. This gives you an idea of the theory of regression analysis, the methods used in computation, and an interpretation of the results. The next reading applies regression analysis to the guns vs. butter trade-off. Expanding this analysis requires factor and discriminant analysis which is described in the predicting expropriation in Jamaica article. The final analysis extends the guns vs. butter example using regression, factor and discriminate analysis

**Presentation**

**Main Readings**

- When to use regression analysis
- How to interpret r-squared regression
- Checking residual plots in regression analysis
- How high should r-squared be in a regression?
- How to interpret coefficients and p-values in a regression
- Optional: A more formal introduction to regression analysis using SPSS - Just down to section 1.3
- P.C. Frederiksen and R.E. Looney, "Defense Expenditures and Economic Growth in Less Developed Countries: Some Further Empirical Evidence" (Journal of Economic Development, July 1982).
- Randall Jones, A Model for Predicting Exropriation in Latin America Applied to Jamaica, Colombia Journal of World Business, Spring 1980.
- Robert Looney and P.C. Frederiksen, "Defense Expenditures, External Public Debt and Growth in Developing Countries (Journal of Peace Research, vol. 23, no.4, 1986)

**Supplemental Readings**