The description of the variables makes it clear that several variables can have country biases by their very construction.
There is a strong correlation between the country effects in the various equations and the various variables.
At the same time, it avoids the country-specific errors associated with the construction of the data which are eliminated under the within-country estimation.