And often just stupid:
The economist Isaac Ehrlich, in an oft-cited 1975 paper, put forth an estimate that is generally considered optimistic: executing 1 criminal translates into 7 fewer homicides that the criminal might have committed. Now do the math. In 1991, there were 14 executions in the United States; in 2001, there were 66. According to Ehrlich's calculations, those 52 additional executions would have accounted for 364 fewer homicides in 2001 -- not a small drop, to be sure, but less than 4 percent of the actual decrease in homicides that year. (my bolding)While I'm opposed to capital punishment and don't want to dispute this conclusion, the reasoning seems absurd: this guy said this, other people think he was being optimistic, therefore the truth couldn't be greater than what he said it was. Because his estimate is considered optimistic, they use it as the top limit of what could possibly be. I mean, give me at least some reason why it's considered optimistic. The "generally considered" argument is exactly the kind of reasoning this book purports to be against.
Also, comment from my brother, who is a statistician (added to this post on April 27):
52 executions in 1991 don't result in 364 fewer homicides in 2001. They didn't say that the reduction was 7 per criminal PER YEAR, just 7 per criminal (over the course the the criminal's remaining life, maybe?). I can't believe that each of these criminals commits 7 homicides per year on average. Without wanting to use a "the actual figure is more like..." argument, I'm pretty sure the actual number for a single year is much much lower. To know by how much the actual number would have been reduced in 2001, you need to know something about the age distribution of the criminals that were not executed (were they the oldest, the youngest? would they even still be alive in 2001?), the amount of time remaining on their jail sentences (would they still be in jail in 2001?) and the ages of criminals when the commit homocides (maybe the ones they decided not to execute would have been past their killing 'prime' 10 years later).So, my brother points out, the statistic isn't so much optimistic as unrealistic.
Another, more offensive, badly supported passage:
Women's rights advocates ... have hyped the incidence of sexual assault, claiming that one in three American women will in her lifetime be a victim of rape or attempted rape. (The actual figure is more like one in eight -- but the advocates know it would take a callous person to publicly dispute those claims.)Leaving aside the lameness of the self-congratulation there for being so tough as to stand up to women's rights advocates, something no man has ever been brave enough to do before, apparently, umm, the "actual" figure -- how could you possibly know? You maybe could say something like, 'most serious studies have estimated' or 'a better statistical model would lead us to think' or ... but just randomly stating that you are in possession of an actual -- even if approximate -- number of rapes and attempted rapes in a woman's lifetime. Who knows? maybe their statistics are so damn sophisticated that I would be amazed by how precise they can get with these things -- but they're certainly not giving me information to assess that. And nothing in this book gives me confidence in this duo's ability to be objective and rigorous. Maybe the economist guy is rigorous in his economics work, but here -- in this book that's been rightly criticized for not being economics at all, which wouldn't bother me except don't lay claim to expertise as a basis for believing your story if you don't have any expertise in what you're actually writing about -- they're playing fast and loose with the numbers.
I won't go through all of the examples of offensiveness, since there were many. One of them, in which they are apparently not being racist (unlike in other parts of the book, in my opinion), is about job interviews. It's a proven fact that, in the U.S., people with non-whitey names don't get called for interviews as much even if their CVs are as good. They don't argue with that. But, they say, maybe it's not racism but a reasonable analysis on the part of employers that people with these non-whitey names are disproportionately from poor, uneducated parents and children of poor, uneducated parents succeed less in life, and therefore that these candidates are a bad bet. They're saying, it would be wrong to discriminate against candidates because they're black, but it's okay to discriminate against them because they are also from less privileged backgrounds. That's just good statistics on the part of the employers, not hiring poor people, since poor people are dumber, statistically. There's so much wrong with the auxiliary assumptions that go into this statement (like, maybe the reason these people do less well is BECAUSE no-one employs them rather than it being the case that an employer wouldn't hire them because they're likely to be dumb and lazy), but it also totally ignores the fact that the CVs used in the studies are just as good as those of the whitey counterparts who do get called back, strongly indicating that at least these candidates with forsaken names have not fallen into their statistical pit of failure.
There is so much more to complain about but I will stop here -- okay, one last thing: it bothers me so much when they seem to think that they've come up with an amazing question, like, "What do maximum-security prisoners have to do with Dutch tulip farmers?" (that's not one of their questions, but they're all like that). They have one of these questions at the beginning of each chapter, and in one chapter they have a section on how to come up with clever questions like these. The trouble is, they didn't come up with those questions. They came up with facts and then formed a question afterwards, based on what they'd looked at. That's not questioning, that's just using rhetoric to introduce your topic.
The basic problem all over with this book is that it makes conclusions based on small bits of information without properly situating those bits of information in a larger social or historical context. Also that it's self satisfied and not as clever, and certainly not as revolutionary, as it thinks it is.
