Classification and Causal Modeling of Economic Indicators of Instability in Developing Nations.

摘要

This study examined the relationships between several quarterly economic time series and an aggregate stability index for developing nations. The hypothesis is that worsening economic conditions in developing nations may be a driving force behind unstable conditions in work, social, military, or political domains. The objectives were to identify economic time series that may serve as indicators of instability and to develop a predictive methodology of instability given the economic input. Classification of the economic series paralleled the U.S. economic indicators of the business cycle. The series were classified as leading, lagging, coincident or unrelated to the instability index. Graphical and cross-correlation analysis were used to determine the type and strength of these relationships. The causal models used regression, logit, cluster, and factor analysis. Regression analysis using both principal components and relative changes values from the previous period was used to see if a subset of the economic series was statistically significant when regressed against the stability index. Keywords: Reports; Military publications; Periodicals. (eg)