A new initiative - http://www.predict09.eu - has entered the scene making outcome predictions for the upcoming EP elections based on national opinion polls and a statistical prediction model using experiences from past European Parliament elections.
The results they display on their main page differ visibly from the results of a combined opinion poll published in February, so the mathematical model used seems to have an effect on the predictions.
It's political science, so we'll see whether it works or not. If it doesn't work, we'll get good explanations why. That's political science.
Update: A critical account from Belgium.
Under the category "European parliament elections 2009" I am following up national and European activities on the path to the European Parliament elections 2009.
For an overview over all articles in this category have a look at the overview article.
For the five newest post see also the sidebar.
Brexit: breaking out of the inner circle
14 hours ago