My original post on the Hoppe case was already my longest ever, so I am updating it here.
VC reader Gabriel Rossman points me to this article by University of Rochester economist, Steven Landsburg, in Slate, which discusses an NPR story that pointed out that gays in California were 70% more likely to smoke:
I’ve just learned from NPR’s All Things Considered that in California, gay men and lesbians are 70 percent more likely to smoke than the general population. In a sterling example of why I try not to listen to too much NPR, reporter Sarah Varney immediately segued into the perceived need for more anti-smoking ads targeted specifically at gays.
In other words, Varney implicitly assumes that gays are either too stupid to have gotten the message that smoking is bad for you or too irrational to have modified their behavior accordingly. A more inquisitive reporter might instead have raised the obvious question: What good reasons might gays have to smoke more than other people?
In four minutes of air time, the closest Varney came to addressing that question was to suggest that for gays, stepping outside for a cigarette can be a good way to meet people—as if the desire to meet people somehow differentiates gays from straights. At the same time, she managed to overlook the blindingly obvious: Gays are disproportionately childless, and childless people are more likely to smoke.
As a matter of fact, childless households (whether gay or straight) spend, on average, 56 percent more on cigarettes and alcohol than their childbearing neighbors. (Among households where the parents have some education, the discrepancy is even larger.) Nor is there anything mysterious about why. First, parents have extra reasons to live long and stay healthy, both so they can be there when their kids need them and so they can enjoy the company of their grandchildren. Second, parents have extra expenses—starting with diapers and continuing through college tuition—that leave less disposable income for cigarettes. Third, a lot of parents don’t like the idea of smoking in front of their children.
As I stated in my original post, this is the way that economists talk. Some people may find it offensive, but it is completely unremarkable in the discipline.
A check of GSS data on smoking shows borderline significance (1-tailed p=.086) for those having same-sex experience in the last 5 years: 38% of gays and bisexuals smoke compared to 30% of others, so any differences nationally are probably less than in California. If I do a logistic regression, controlling for education and region, then the 1-tailed significance equals the .05 threshold, meaning that (with controls) gays do appear to smoke more in the GSS data. When being gay is measured by the gender of sexual partners since age 18, the effect is not significant.
As to driniking, gays and bisexuals (measured by activity since age 18) are more likely (45%) to report ever getting drunk than others (35%), though the effect is not significant for gays measured by activity in the last 5 years.
As to the evidence in this post, it provides some (but far from conclusive) support for Hoppe’s claims that gays engage in higher risk behavior and that differences in child-rearing is a related cause. So now, four pieces of evidence provide some support Hoppe’s claims: a greater reported number of sex partners (as I linked in my earlier post), a higher rate of smoking in some studies (but not others), a higher rate of getting drunk (by some measures but not others), and some plausible evidence that child-rearing is related to smoking and drinking rates. The direct evidence of less planning for the future is not shown in these studies, but is consistent with some of them. The evidence that I pointed to against Hoppe’s thesis is mostly attitudinal, not behavioral, and economists (though not sociologists) tend to discount attitudes.
The caveats that I mentioned in the prior post apply to my analyses here as well, including the absence of adjusting the sample size for a design effect. I’m off to see a movie; I’ll probably post more when I get back today–or post tomorrow.
UPDATE:
For more, see David Beito, Ralph Luker, Kenneth Gregg, and Tom G. Palmer. Palmer, a Senior Fellow at the Cato Institute, considers Hoppe a bigot, not for the statements attributed to Hoppe in class (which he finds unobjectionable), but primarily for other comments made elsewhere about Palmer. Nonetheless, Palmer has written to UNLV in support of Hoppe’s academic freedom.
Comments are closed.