This article took me back. About a year, to be more specific. A noted Berkeley sociology professor, Michael Hout, and three graduate students posted a paper alleging that there was a serious anomaly in the 2004 presidential vote in three Florida counties, Palm Beach, Broward, and Dade. The paper claimed that Bush got "too many" votes in those three counties and suggested that fraud was possible. This alarming conclusion was widely publicized.
I noted that the study omitted an important characteristic of those three counties, obvious to anyone who knew anything about them, a relatively high proportion of Jewish residents. I suggested that Jewish voters might have voted in unusual numbers for Bush because 1) Bush changed our policy toward Yassir Arafat, and toward the Middle East generally, in ways an awful lot of Jewish voters approved of, and 2) an unknown, but anecdotally high, proportion of South Florida Jews are originally from the New York City metro area, and 9/11 could well have been exceptionally important to them. I did some statistical work that indicated this could, in fact, account for the "anomaly". (And see a follow up.)
What does the article have to do with this? I used data for Jewish population by county that were crude but the best I could find quickly. That data stated Palm Beach County's Jewish population was 167,000. The article, from the Palm Beach Post, reports that a new study of Palm Beach County's population is expected to report a Jewish population of 254,300. 20% of the county's population is Jewish, the highest for any major geographic area outside of Israel.
I think it is thus even more likely that the Hout, et al. results were due to their failure to consider Jewish voters.