Some indicators of governance quality capture the quality of inputs to the ‘production’
of governance and others reflect government output; some operate at the underlying level
of fundamentals and others reflect direct economic impact. However, all of them are
interrelated. For that reason, we treat them separately in the empirical analysis, including
one dimension of governance in the equation at a time. Adding too many at once results in
serious problems of multicollinearity in the regression model.
4. Basic Results
Before investigating the effects of institutions, we first discuss the regression results for
a set of specifications of the gravity equation that take into account the standard variables
often applied in the literature. They include gross domestic product (GDP) for both the
country of origin and destination (‘home’ and ‘foreign’), geographical distance and several
variables that have proven to be effective controls for shared historical, political and
cultural background (see Frankel et al., 1997). The latter are dummy variables that indicate
the presence of a common border (adjacency), common language, common dominant
religion and common colonial history. Furthermore, we control for the effect of economic
integration policies by means of a dummy variable: common trade block (often labelled
‘free trade agreement’). Table 2 presents an overview of the results of running OLS
regressions on six differently specified basic versions of the gravity equation.
Bilateral Trade Flows and Institutions