The data set used in the estimation is a panel data set in two dimensions: the trading
partner dimension and the time dimension. In panel data, pooled ordinary least squares
(OLS), fixed-effects (FE) and random-effects (RE) estimators are alternatively used in
this type of study. However, due to small sample problems, by which the limited
number of cross-sectional units and the limited number of available data over time led
to the violation of basic assumption of standard statistical analysis, pooled OLS
estimator is applied in our study. Advantage of pooling is to increase the sample size,
thereby obtaining more precise estimates and test statistics with greater power.
Moreover, pooled OLS analysis concerns the possibility to capture not only the
variation of what emerges through time or space, but the variation of these two
dimensions simultaneously. This is because, instead of testing a cross-section model
for all countries at one point in time or testing a time series model for one country using
time series data, a pooled model is tested for all countries through time (Podesta, 2000).
We also assume that there are no significant country (individual) effects. In the context
of this study, it means that all of Korea’s trading partners in the sample would react in