Both graphs show the BIP. The difference between the two? The upper one shows the quarterly numbers in billion €, the lower one only the annual numbers. So far with me?
Alright, then. Looks easy enough to read, right? We have our dip in 2009, slowly climbing back up, and the final number on the lower graph is that of 2008. So let us compare the 2008 numbers only, shall we?
The 2008 quartely numbers are
Q1 = 609,70
Q2 = 625,60
Q3 = 634,40
Q4 = 626,10
The annual cumulative number shown in the lower graph for the BIP is indicates as
2008 = 2492,00
Do a bit of math here with me, will you? It's easy math. it's called addition.
All four quartals of 2008 as shown in the upper graph should equal the cumulative number of the lower graph for 2008, right?
Or, in mathematical terms
G1(2008) = G1( [Q1 + Q2 + Q3 + Q4] 2008) = G2(2008)
609,7 + 625,6 + 634,4 + 626,1 = 2492,0
2495,8 = 2492,0
We just lost 3.8 billion. But hey, who's counting? At Spiegel (which I usually consider to be a good publication) in this instance, apparently nobody.
I mean, sure, these 3.8 billion only come up to 0,15 percent of "more growth", but at a time when politicians would suck cock for any and all 0,x percent of "more growth", one would think that somebody might have noticed such an error in adding up simple numbers.
If one does roughly 30 seconds of searching, one can find the actual published numbers by the Statistische Bundesamt (sourced here in the graphs), with the numbers for for 2008.
Here (you need to scroll a bit) you will find the actual numbers that add up to the aforementioned 2492,0 billion €.
However, the source numbers are not the ones they use in their publications. Nope. Those numbers are cleaned up based on a model called Census X-12 ARIMA., which gets you entirely different numbers once more.
And we don't even want to get into these things, because that way lies madness.
Suffice to say, though, in PR releases and articles based on these, there often happens a glorious mix'n'match between different models...
You can do whatever you want them to do to prove whatever point you want to make.
Provided you are intelligent enough to obscure things.
Maybe journalists should start to at least try to learn the methodology. If only to be able to make the distinction between what may be right and what is absolute bullshit.
Nah. Too much to ask.