Another Sunday, Another Sonntagsfrage

It's Sunday and, thus, it seems appropriate to delve into another Germany political tradition--the Sunday question (Sonntagsfrage): "If the Bundestag election were next Sunday, for which party would you vote?" All of the most recent results continue to see a bare majority (51%) for a CDU/CSU-FDP coalition, mainly because the FDP is up by 2% (according to ARD Deutschlandtrend, http://www.tagesthemen.de/inland/deutschlandtrend/index.html). The SPD continues to hemorrhage support, now down to 23% (or even 22% according to some sources). The train wreck that is the SPD this year is simultaneously sad and entertaining to watch--but that is a topic for another day.

Just like in the U.S., polling has become a veritable obsession in German election campaigns with numerous companies and organizations providing their findings for the German electorate and pundits to mull over. But also like the U.S.--remember the exit polling debacle from 2000--German pollsters were rather imprecise the last time around. All Sonntagsfragen predicted that the CDU/CSU would poll about 40-41% on election day--instead they garnered only 35%, barely more than the SPD. That is a pretty significant overestimation, outside of the commonly accepted margin of error range (2-3%). Academics and pundits have been speculating about this ever since and, unsurprisingly, many are questioning the accuracy of the polls this time around.

So, what happened in 2005 that might or might not happen again this year? One potential problem is the accuracy of samples--with the propensity of older people to respond more frequently than younger folks (and thus over-estimating the support of the CDU). Another growing problem in the US (and probably Germany) given the dependence on landlines by polling organizations, is the rise in cell phone-only people (http://thecaucus.blogs.nytimes.com/2008/07/23/cellphone-only-it-holds-little-sway-in-polls/) who are generally much younger than average. It is possible that this segment of the German electorate was under-sampled before the 2005 election (and this year?).

The most common explanation for the polling problems in 2005 was the terrible campaign that the CDU waged--the whole Kirchhof debacle (announced as the next finance minister but having advocated extremely unpopular flat taxes). Basically, the consensus is that Merkel and the CDU campaigned too far to the right and scared many voters away with unpalatable neoliberal reforms at the last minute. It should be noted that despite 15 years of party and public service, she was considered to be a campaign novice and relatively untested. She is a much more savvy and experienced politician in 2009.

We should also not forget that the SPD was still led by Gerhard Schröder in 2005--like Bill Clinton, probably the most charismatic politician of his generation. Despite everything, he was able to increase the SPD's share of the vote by 1% over pre-election polling prognoses. We should not forget how he "came from behind" also in 2002 to win the election. As late as August 2002, the CDU/CSU was winning 40% to 34%--and the actual result was a tie at 38.5%. (http://www.aillyacum.de/Dt/Wahlen-Deutschland/2002/index07.html) That year, Schröder was able to use the floods of the Elbe river in the east to his party's advantage (as well as splits in the CDU/CSU between CDU party leader Merkel and CSU leader and Chancellor Candidate Edmund Stoiber)--and in 2005 he mercilessly attacked the CDU over the Kirchhof/tax saga and other issues. Obviously, the absence of a leader like him with such shrewd instincts and communicative skills will hurt the SPD this time around.

Schröder brings up two final thoughts. First, eastern German voters have been extremely volatile since 1990. Predicting the eventual outcome there--becoming the key "swing" states with every eventual chancellor since 1990 winning there--is especially difficult, but highly important. Clearly, Schröder's use of the floods in the east and then his attack on flax tax proposals in 2005 were meant to curry favor in that perpetually depressed region. In any case, especial scrutiny of campaign dynamics in this region is always warranted.

Finally, despite all of the talk about how boring this election season is, the campaign with all of its vicissitudes, scandals, missteps, etc., does matter, and like 2002 and 2005 may very well be decisive on September 27th--regardless of what the Sonntagsfrage indicates.

--Eric Langenbacher