Welcome to European Tribune. It's gone a bit quiet around here these days, but it's still going.
First, let me welcome you to European Tribune

You claim, in your linked post, that using a fixed-effect estimator suffices to account for the effect of being a member of the Eurozone.

This is obviously false. In order to properly account for higher-order effects, you would have needed cross-terms of a Eurozone dummy with each of your independent variables. Just as (and for the same reason) you include cross-terms between the independent variables in question.

As you correctly note, this would have reduced the statistical power of your test. You are incorrect about the magnitude of the reduction, however: Including Eurozone dummy cross-terms would increase the number of estimated coefficients by only five, not one hundred as you allege.

But more fundamentally, the fact that your data is insufficient to power a proper estimator does not in any way, shape or form justify using an obvious misspecification. In point of fact, using an obvious misspecification to generate fake statistical power is a strong indicator of pseudoscience in action, and hence an argument in favor of the original diary's conclusion.

- Jake

Friends come and go. Enemies accumulate.

by JakeS (JangoSierra 'at' gmail 'dot' com) on Thu Mar 14th, 2013 at 06:01:51 PM EST
[ Parent ]

Others have rated this comment as follows:


Occasional Series