You cant really compare unemployment numbers from the US to those of the European countries.
The US number is a poll, a scientific poll but a poll, where as most European numbers are actually very exact numbers in comparison.
In most European countries, unemployment numbers are based on people applying for unemployment (aka registering for it) payment, which gives a very accurate number, and you can stay registered forever or until you are thrown over in another statistical pool (not employable, early retirement or whatever the country uses). You do not fall out of the system just because you have been unemployed for 3 years for example
As I understand the US number based on the official website, the US number is a statical model with information gathered from many different sources, that may or may not be accurate. On top of that people who run out of unemployment insurance, simply drop off the radar. This warps the actual unemployment number (lowers it) considerably. For example, thanks to this and other factors the official unemployment number has either fallen or stayed the same despite there being tons of job losses.
With a group of 100 professional men and women (non-farm labor,) working for a variety of different corporations, at the first of month No. 1, 4 percent of these professionals are cut. With 96 people employed, 4 people are on unemployment benefits for the next 10 weeks. And our unemployment rate is 4 percent.
At the first of month No.2, two more people are laid off... six people are on unemployment benefits for the next 10 weeks, and six weeks respectively. While 94 people are employed the unemployment rate has risen to 6 percent.
Month No.3, four of our original 100 lose their benefits, so, 96 people are counted in the pool. Three people are laid off and our total of unemployed professionals stands at five, but the unemployment rate has dropped from month No.2, to 5 percent without anyone having found a new job.
Month No.4; great news two people have found jobs (one from month No.2,) two people are laid off, but we have a second readjustment to the number of employed, because only the two hired now count. With 95 people in the pool, our unemployment rate holds steady at 5 percent.
Month No.5; more great news two people found jobs! One person fell off the benefits, three people are laid off and now, we only have 94 people to measure. The unemployment rate holds steady at 5 percent.
Month No. 6, no new jobs this month with our 94 people. Two people were laid off. We loose two people on benefits. We have five people on benefits in total, and now with only 92 people to measure, our unemployment rates holds at 5 percent.
Month No. 7; excellent news two people were hired this month! Three were laid off. Four people live on benefits and one person fell off. And our unemployment rate has fallen to 4 percent while the number of people on benefits is at the lowest in six months.
Taken from this old article, but it is a good explanation.
T&A: Unemployment Rate United States
That is why the true US unemployment number is far higher than the actual.
On the other hand, European countries (like the US also does) love to "hide" unemployed people in different statistics. For example, in Denmark when a person goes for retraining, he moves from the unemployment statistics to the education statistics... yet he still receives unemployment insurance or similar.
Also the way that countries handle part time employees is different even with-in the EU. In Denmark part time is around 15 hours a week for example, and are only counted as half a full time job, where in the US I believe half time employment is counted as full time.
All this, and more, makes unemployment comparison near impossible between the US and EU, and even with in the EU to be honest.