Having gauged separately the psychological responses of our sample to the Madrid attacks, we created a summary measure for each respondent by adding across the fifteen symptoms in our inventory. Nationally, 41 percent reported experiencing at least one of these symptoms, compared to 57 percent in Madrid.

We divided the samples into two parts: those who stated that they had experienced none of the fifteen symptoms and those who said that they had experienced at least one. When vote choice was examined by reported symptoms comparing these two groups, no difference in support for the PP was found for the national sample (Figures 4A and 4B).

In Madrid, however, a substantially smaller percentage of those who said that they had experienced at least one symptom reported having voted for the PP (Figures 5A and 5B).

We mentioned earlier that about a third of the country thought that Spain might have done something to bring on the attacks. Although we did not ask what they thought Spain might have done, it is highly likely that citizens were thinking about their country’s involvement in the war in Iraq. To explore this further, we again divided the sample into two, separating out those who put virtually no stock in the idea that Spain had brought the attacks on through its own actions (responding “never crossed my mind” or “not very deep concern”) from those who were convinced of this idea (deep concern/one of my deepest concerns), and examined reported vote choice for the national sample and for citizens of Madrid. At both the national and local levels, a lower percentage of citizens who held Spain responsible for the attacks reported voting for the PP compared to those who did not hold Spain responsible (Figures 6A, 6B, 7A, and 7B). The difference in support was nearly three percentage points less among those in Madrid compared to those in the Spanish national sample.