A Public Opinion Riddle of Sorts

I have spent the past couple days toying with my introduction. Earlier in my research, and even through most of the middle stages, I definitely sidelined the planning for my intro to focus on devising a plan for data analysis in the statistics program I’m using, called R. I had reasoned that learning how to code in R would be a a handful, and would take up most of my time. I was not wrong, since I just spent several days figuring out how to find the perfect x and y coordinates to position my legend in a proper spot in my bar plots. But the trials and tribulations of  data analysis in R are for another day. For now, I would like to present a short narrative on the origins of the theory my study is testing. I developed the narrative throughout my literature review — it reads like a public opinion riddle (of sorts). See if you can solve it before my next update:

During the 1965 German federal election, Dr. Elisabeth Noelle-Neumann and her team at the Allensbach Institute for Public Opinion Research noted two puzzling trends in public opinion among German voters.

The first trend was a sudden shift in voters’ intentions, reflected in a final survey. The two major political parties, the incumbent Christian Democratic Party and the opposing Social Democratic Party, had been locked in bitter stalemate for the past ten months of the campaign, with each party consistently polling at 45% of the vote. In the weeks leading up to the election, however, the Christian Democrats’ predicted vote share swiftly climbed to 50%, while the Social Democrats’ share dropped below 40% (Kaid 2007).

The second trend was the absence of a direct correlation between voters’ intentions and their predictions in all prior surveys. While the proportion of voters who reported that they would vote for either the Christian Democrats or the Social Democrats if the election were held that week remained constant for both parties, the proportion of voters who expected a Christian Democratic victory steadily increased to 50% from December 1964 to August 1965. This final survey result matched the election outcome: the Christian Democrats won the election with 49% of the vote, while the Social Democrats received only 40% (Kaid 2007).     

The pollsters at the Allensbach Institute thus wondered: How  could voters’ expectation of a Christian Democratic victory have increased steadily while voters’ reported intentions come election time remained constant, and when the survey results consistently predicted such a close race (Kennamer 1990)? 

I am considering using this riddle, after I develop it further a bit, to begin my paper. At first, I considered beginning my introduction by funneling down from deliberative democratic theory, rather than funneling up from an empirical puzzle. But  I’ve always had a soft spot for empirical observations as hooks for research studies. I think I’ll stick with the riddle.


Kaid, L. L., & Holtz-Bacha, C. (Eds.). (2007). Encyclopedia of political communication. SAGE publications.

Kennamer, J. D. (1990). Self-Serving Biases in Perceiving the Opinions of Others Implications for the Spiral of Silence. Communication Research, 17(3), 393-404.