wordpress-themes.org wordpress themes wordpress themes

w6707-27
A Second Step: Conditioning on Country Characteristics

We now present estimates from our complete model allowing for the interaction of country characteristics with a measure of openness. We report only interactions with the trade intensity measure of openness because other candidate measures have either very little time series variation or were eliminated because of insufficient data. We report the results for our second step procedure for finding trade’s effect in table 3 fully.

Note: T-statistics are shown in parentheses. Significance at the 95% and 99% confidence levels are indicated by * and **, respectively. Dependent variable is the log of the median of SO concentrations at each observation site.

There are several features of note in the table. First, adding the openness interactions has not undermined the model’s basic predictions regarding scale, technique and composition effects. In particular, the sign of our basic regressors is maintained and in most cases the significance levels are enhanced by the inclusion of the openness interactions.

Second, the inclusion of country characteristics appears to have made a large difference to the impact openness has on pollution. The coefficient on our measure of openness is now highly significant whereas in table 2 it was not significant at conventional levels. Its magnitude is now approximately ten times its former size. The interaction terms with country characteristics are also highly significant.

Third, the sign pattern of the interaction terms is as expected from theory. The linear interaction term on openness and (relative) income per capita is positive in both columns and the quadratic term is always negative. Consequently if a country has a relatively low level of income per capita relative to the rest of the world, then – all else equal – the impact of further openness must be to make this country dirtier. Relatively rich countries would be made cleaner with trade. These results may reflect the ceteris paribus pollution haven hypothesis where relative income differences alone determine the composition effect of trade.