Statisticians have long warned that correlation is not causation, but they have apparently warned in vain.
There is no reason to doubt that heavily armed police in riot gear may be more likely to show up where outbreaks of violence are expected. But when violence then breaks out, does that prove that it was the appearance of the police that caused it?
I strongly suspect that people who travel with armed guards are more likely to be murdered than people who do not travel with armed guards. After all, they are not paying to have armed guards for no reason.
If so, should we conclude from a higher murder rate among people with armed guards that having armed guards increases your chances of getting murdered? Shall we also conclude from this that we the taxpayers should no longer pay to have Secret Service agents guarding our presidents?
Actually, the history of assassinations of American presidents could be cited as evidence that armed guards are correlated with higher murder rates, if we proceed to "reason" the same way the advocates of weaker police presence seem to be reasoning.
There have been 43 Presidents of the United States, of whom four -- Lincoln, Garfield, McKinley and Kennedy -- have been murdered. That is a murder rate of 9 percent.
If the murder rate in the general population -- most of whom do not have armed guards -- were 9 percent, that would mean more than 27 million Americans murdered today. We haven't quite gotten up to a murder rate that high, even in Chicago.
Does anyone seriously believe that leaving presidents unguarded would reduce assassinations? Probably not. But this is the golden age of talking points, as distinguished from serious thinking about serious issues.
These talking points are often based on a prevailing social vision, rather than on hard facts. According to the prevailing vision, ghetto riots are due to racial injustices -- and the way to deal with them is to make concessions in words and deeds, while severely restricting the use of force by the police.
Factual evidence cannot make a dent in that vision.
But, for those who are still so old-fashioned as to rely on facts, here are a few: Back in the 1960s when ghetto riots broke out in cities across the country, the region with the fewest riots was the South, where racial discrimination was greatest and police forces least likely to show restraint.