I find this extremely difficult to believe. Now, perhaps 10% of Europeans are practicing Christians. But just think about Poland. Ireland, Spain, Italy. I sincerely believe if you asked the people in those countries whether they considered themselves to be Christian or not, an overwhelming majority would likely say yes. The CIA factbook lists 40 Million Christians of one type or another in Great Britain alone. That's at least ten percent of Europe right there. In Bavaria, they have the crucifix in the classrooms and it seemed to me that most Germans go through baptism and confirmation ceremonies.
Again, practicing- that's another matter. But 10% is hard for me to believe.
Survey costruction is notoriously difficuly because, among other reasons, the wording, format, and sequencing of the questions affect the answers. Indeed, when asked in this form, many more than 10% answer that they are Christians.
No offense intended by the following comparison, but consider asking people whether they are good lovers, say. What do you think the outcome would be? A great over-reporting; many more people would self-report themselved to be good lovers than actually are there.
What do the two cases have in common? Essentially this: when there is a popular or ethically dictated answer to the question, that answer will be given, in many cases falsely. Thus, the desire to be viewed as successful with the opposite sex leads to overstimation in the second case. The view, adopted from parents in childhood, makes people consider being Christian as synonimous with being a "good person." This too leads to overestimation of one's adherence to, and not just the practice of, his family's religion.
In both of the aforementioned cases, if one want a better handle on reality, one has to rely on better measures than self-reporting, which is clearly flawed. To me, attendance of Churches and synagogues is a much better --- although not ideal, naturally --- measure of one's beliefs.
It is in that sense that I have used the word "identify:" those attending churches identify by their behaviors themseves as Christians.