Election campaign as horse race
Column: Whether Germany’s next chancellor is named Merkel or Schröder, this election was certainly a lesson in how the media exploit opinion polls.
Column Whether Germany's next chancellor will be Merkel
or Schröder, this election was a lesson in how the media instrumentalize the polls.
Germany made history. In the end, this time it was actually a
Head-to-head race between the two shrunken mainstream parties CDU/CSU and
SPD. And then, when everything was over, we were allowed to go to the
Elephant round of television a still incumbent Federal Chancellor
who, with a showy arrogance, is able to influence the media and opinion polls.
attacked him in equal measure; as if they were to blame for his defeat.
As absurd as this spectacle was, it provided the occasion for a
Review.
In addition to the media staging of the election campaign, which is now
economic logic in the competition for audience attention,
also includes the role that opinion research plays in the
media coverage. Over weeks and months
there are the latest "water level reports" almost daily on the so
Sunday question (How would you vote if next year's
weekend?). Journalists play the
different institutions against each other, compartmentalize and
over-interpret divergent election forecasts - however
usually without researching which researchers work how seriously and
who have been known to cheat in order to keep their clients happy.
enjoy. There is also little discussion about the influence of the
surveys have on actual behavior: After all, this time
a few percent more voters would have been smart enough to vote with
strategically base their voting decision on the forecasts.
the CDU/CSU, for example, from becoming too powerful.
let. In this way, opinion research becomes "self-fulfilling" wherever possible
or the "self-destroying prophecy".
Likewise, many journalists stubbornly refuse to acknowledge
that election predictions cannot be accurate. They indicate
fluctuation margins and therefore inaccuracies. Dispute over deviating
percentages rather reveals the - striking! - ignorance of many
policy editors with regard to the methodology of empirical
survey research rather than being enlightening on the matter itself.
However: If media representatives were to inform their audiences about the limited
inform about the informative value of election polls, they would also be
favorite toy during the pre-election phase. And thus one
cost-effective way to dramatize the election as a "horse race"
and thus increase circulation and ratings. This time the deviations were
between the latest estimates of the pollsters and the election result
However, the differences are so great that even the pollsters themselves are puzzled.
The grande dame of demoscopy, Elisabeth Noelle, as well as the
renowned American professor of journalism Philip Meyer already
decades to promote "precision journalism". Both
call for more social science expertise in the editorial offices
handling survey data and statistics.
There are hardly any media-related degree programs today in which
such basic knowledge would not be taught. But it has borne fruit
hardly ever. Surveys and other social research are then
reported if they have entertainment value. How and why the
Unfortunately, too few people are interested in the results
Journalists. And also who the client of which
study, they all too often withhold them from their audiences.
or Schröder, this election was a lesson in how the media instrumentalize the polls.
Germany made history. In the end, this time it was actually a
Head-to-head race between the two shrunken mainstream parties CDU/CSU and
SPD. And then, when everything was over, we were allowed to go to the
Elephant round of television a still incumbent Federal Chancellor
who, with a showy arrogance, is able to influence the media and opinion polls.
attacked him in equal measure; as if they were to blame for his defeat.
As absurd as this spectacle was, it provided the occasion for a
Review.
In addition to the media staging of the election campaign, which is now
economic logic in the competition for audience attention,
also includes the role that opinion research plays in the
media coverage. Over weeks and months
there are the latest "water level reports" almost daily on the so
Sunday question (How would you vote if next year's
weekend?). Journalists play the
different institutions against each other, compartmentalize and
over-interpret divergent election forecasts - however
usually without researching which researchers work how seriously and
who have been known to cheat in order to keep their clients happy.
enjoy. There is also little discussion about the influence of the
surveys have on actual behavior: After all, this time
a few percent more voters would have been smart enough to vote with
strategically base their voting decision on the forecasts.
the CDU/CSU, for example, from becoming too powerful.
let. In this way, opinion research becomes "self-fulfilling" wherever possible
or the "self-destroying prophecy".
Likewise, many journalists stubbornly refuse to acknowledge
that election predictions cannot be accurate. They indicate
fluctuation margins and therefore inaccuracies. Dispute over deviating
percentages rather reveals the - striking! - ignorance of many
policy editors with regard to the methodology of empirical
survey research rather than being enlightening on the matter itself.
However: If media representatives were to inform their audiences about the limited
inform about the informative value of election polls, they would also be
favorite toy during the pre-election phase. And thus one
cost-effective way to dramatize the election as a "horse race"
and thus increase circulation and ratings. This time the deviations were
between the latest estimates of the pollsters and the election result
However, the differences are so great that even the pollsters themselves are puzzled.
The grande dame of demoscopy, Elisabeth Noelle, as well as the
renowned American professor of journalism Philip Meyer already
decades to promote "precision journalism". Both
call for more social science expertise in the editorial offices
handling survey data and statistics.
There are hardly any media-related degree programs today in which
such basic knowledge would not be taught. But it has borne fruit
hardly ever. Surveys and other social research are then
reported if they have entertainment value. How and why the
Unfortunately, too few people are interested in the results
Journalists. And also who the client of which
study, they all too often withhold them from their audiences.
> Stephan Russ-Mohl is a professor of journalism at the Università della Svizzera Italiana in Lugano.
