The first explanation seems unlikely. If the assumption that the strength-of-personality scale can identify opinion leaders is fundamentally wrong, why did those identified via the scale display the traits that had been ascribed to opinion leaders for decades before 2002, and why were there so many cases where they did, in fact, set the trend and were subsequently followed by the rest of the population?

There is also little evidence that the strength-of-personality scale had lost its validity. First of all, it hardly seems plausible that the scale could have degenerated within only half a year’s time, since the turnabout in favor of the Christian Democrats in early 2002 followed the familiar pattern, with opinion leaders preceding the remaining population in the opinion formation process. Second, surveys conducted in the final stage of the campaign showed that those identified as opinion leaders were still more interested in politics, better informed, and politically more active than the remaining population, just as they had been in prior years. Moreover, the share of personal persuasion attempts and, hence, we can assume, the level of activity among opinion leaders, was also not substantially lower than it was in 1998, and was even somewhat higher than in 1994 (Figure 5).

As for the third possible explanation for why opinion leaders relinquished their leading role in 2002, the important question is whether there was anything that fundamentally distinguished that campaign from those waged in prior years. And, indeed, there was: The 2002 federal election represented the first time in the history of the Federal Republic of Germany that a party—in this case, the Social Democrats—managed to win an election even though the population believed right before election day that it was less competent on balance than its opponent to handle various issues in classical political areas such as labor, domestic, and economic policy.

This fact, which was at least as unusual as the opinion leaders’ failure to play any trendsetting role in the final phase of the campaign, indicates that some factor aside from the customary themes must have had a decisive impact on how broad segments of the population decided to vote. Evidence indicates that that factor was a natural catastrophe—a disastrous flood in Eastern Germany, and the media reporting on the events surrounding it.

In mid-August 2002, after weeks of heavy rainfall, Germany’s second-largest river, the Elbe, burst its banks, flooding the town of Grimma and parts of Dresden, the capital of Saxony, among other places. Just a few days later, the topics of conversation among the population changed. In a representative survey conducted by the Allensbach Institute in early September, 92 percent of all respondents said they had talked recently about the flood with friends or acquaintances. Political issues trailed far behind as topics of discussion: 60 percent had talked about the federal election in general, 65 percent about unemployment, and 42 percent the economic situation. In other words, the flood in Eastern Germany had eclipsed all of the other issues that had dominated the campaign up to that point.

The notion that there was a meaningful connection between the new focus on the topic of the flood in the public debate and the improvement that occurred at this time in the Social Democrats’ election prospects is supported by a late August 2002 survey by the Allensbach Institute, in which respondents who said they had followed the reporting on the flood “very closely” also declared an intention to vote for the Social Democrats in the upcoming election considerably more often than those who had not. In contrast, these two groups did not differ with respect to the party they had voted for in the prior federal election (see Table 1).