Today I was browsing through the history of the cold war, and that led me to the realization that the only two senators that have been assassinated in US history were both left-wing populists. That got me wondering if more broadly speaking, if there were any differences in the patterns of assassinations between parties. So I stumbled across this list of assassinations on Wikipedia, copied it into Excel, built a pivot table, calculated 95% confidence intervals about the frequencies, and created this graph.
Of course this is Wikipedia, so the results should be taken with a grain of salt, but for some reason the Republican party appears to have experienced a marked decrease in successful assassinations made against their members in the 20th century while the rate of assassination of Democratic party members has remained statistically unchanged.
What do you think accounts for this disparity? Is Wikipedia biased in its reporting of assassinations? If so, why is that pattern not observed in the 19th century? Is there something about the Republican party in the 20th century that makes it better equipped to rebuff assassination attempts?
Any thoughts you have about what might explain the disparity would be appreciated.
You've made an interesting observation here, but I think this basically luck. Most attempts to assassinate U.S. political leaders are non-political in nature. I'm not sure how you came up with your confidence intervals on a count variable, but I'll also emphasize that assassinations (as defined by the Wikipedia page) are extremely rare. I'm not sure you can draw statistical conclusions under the circumstances.
In 1999, the Secret Service published a study based on analyzing the 83 people who "attacked, or approached to attack, a prominent official in the United States" from 1949-1996 using all of the available law enforcement records, other data, and interviews with the individuals when possible. The study finds:
Attackers and near-lethal approachers of public officials rarely had "political" motives. Only one subject who acted alone (Sirhan Sirhan) might be seen to have a primary political motive or have a primary interest in changing particular government policies (And even in Sirhan's case, there is considerable evidence to suggest that his primary interest in assassinating Senator Robert F. Kennedy was to achieve notoriety.)
Given that political motives rarely animate assassins in the United States, it seems unlikely that the correlation you've found is causal in nature. It's also important to note just how rare they are. While assassination attempts against U.S. presidents are reasonably common, completed assassinations against political figures as defined by that Wikipedia list are vanishingly rare.
The Wikipedia list covers (and surely non-comprehensive) candidates and office holders at every level: state, federal, and local. I don't know of a firm figure on the number of people ever to hold or run for office in the U.S., but it's definitely in the millions if not the tens of millions. There are over half a million elected government positions in the U.S. today. Exactly how many people have held or run for those positions historically is hard to say, but we're clearly talking at least about millions of people. If we call if 6 million, which seems fair, then we're talking about 1 politician in 100,000 being assassinated. That's just insanely, insanely rare. I'd be very reluctant to try to find a pattern there.
And assassination is an activity where a lot depends on luck. Take the case of U.S. Presidents. One U.S. President, a Democrat (JFK), was assassinated in the 20th century, but there were relatively serious assassination attempts against many of his Republican counterparts. Reagan was shot. Ford was nearly shot on two separate occasions. Teddy Roosevelt (albeit running as a progressive candidate at the time) was shot -- and allegedly saved only by a thick speech in his pocket. It's really just chance that JFK and not any of the others died.