Letters from our readers.
Jan. 16 1998 3:30 AM

Address your e-mail to the editors to letters@slate.com. Please include your address and daytime phone number (for confirmation only).


Your Days Are Numbered (Varyingly)

William Bennett has been publicizing our finding about the life span of homosexuals: that following the homosexual lifestyle appears to take a 20-to-30-year bite out of your life span. Walter Olson ("William Bennett, Gays, and the Truth") doesn't like our findings.

So what's new? I've been down this road with the "skeptics" before. In 1967 when I published the first study on the harmful effects of secondhand tobacco smoke, there were a lot of skeptics too. But, over time, my findings have held up rather well. I fully expect to be similarly vindicated on the homosexual-life-span issue. And, like so many on the left, Olson doesn't get his facts straight.

We published our finding in Omega, a peer-reviewed scientific journal. That is, a panel of scientists in the field went over our work with a fine-toothed comb before it got published. As with any findings, they are subject to revision as new information comes to light. But the proper place to criticize them is in another peer-reviewed scientific journal. Olson's method is less than what one might expect from a "senior fellow" at a think tank. Did Olson have his experts read the paper and then record their comments to it? No, Olson verbally sketched our method to various professionals and got their "negative" reactions. The perspicacity of professionals who will go on the record with so little information is subject to question.



Because we investigated several things in our study. Yes, we reviewed 6,737 obituaries. And old homosexuals were conspicuous by their absence. But we noted that in over a decade of obituaries the "numbers didn't change"--that is, the distribution of ages at death stayed essentially constant. That finding was curious, because in a growing movement a decade is a fairly long time. We were well aware that those who appeared in the obituaries met the criteria of Dr. Richard Isay, chairman of the American Psychiatric Association's Committee on Gay, Lesbian, and Bisexual Issues, who contends that gays who are "out" are the healthiest homosexuals of all. But Isay might be wrong. Perhaps the "unknown" homosexuals, those still "hidden" from public view, were the healthiest.

So we checked back through time--all the way back to 1858--and looked at the scientific literature on homosexuals. Some of it was collected for forensic purposes, some for sociological or sexological purposes, some for medical reasons. With almost no exceptions, the pattern in the previous studies of homosexuals was consonant with what we found in the obituaries. That is, the median age of the sample of homosexuals was in the 30s and very few were recorded as old (i.e., age 65 or older). This finding went some distance toward satisfying the objection that what we were sampling in obits was an atypical set of gays and also left about the right amount of "room" for gays to die at a median age somewhere in their 40s. But, like any empirical effort, it did not completely satisfy the objection that we and other investigators hadn't sampled gays "who end their days in retired obscurity in some sunny clime."

Since the publication of our findings, two large random surveys of sexuality appeared--and both reported results consonant with ours. The University of Chicago "definitive" sex survey of Americans reported that 2.9 percent of men aged 18-29 claimed that they were bisexual or homosexual. For those aged 30-39, 4.2 percent made the same claim. But for those aged 40-49 the proportion declined to 2.2 percent, and for those 50-59 it declined to 0.5 percent. The corresponding proportions for women were 1.6 percent, 1.8 percent, 1.3 percent, and 0.4 percent. A British survey that made similar claims to representativeness found that, overall, 1.5 percent of the men and 0.7 percent of the women reported homosexual sex in the past 5 years. But the proportion of those aged 45-59 who made this claim was only 0.9 percent for the men and 0.1 percent for the women. More "missing" gays and more "missing" lesbians.


Olson says that "Harry Rosenberg, the mortality-statistics chief at the National Center for Health Statistics, says he's unaware of evidence that HIV-negative gays have a lower life expectancy than other males. Rosenberg also points to one reason to think the HIV-negative gay male may actually live longer on average than the straight male: Gays may have higher incomes and more education than straights--two factors powerfully correlated with longer life spans."

When I first published my findings about secondhand tobacco smoke, everyone who hadn't read the article could have reasonably said that they were "unaware of evidence" that it was injurious to the nonsmoker. Being "unaware" of evidence is commonplace. Further, speculating on what might actually be the case is something that scientists do all the time. But speculations are quite different from empirical testing. I'll wager that Rosenberg was also "unaware" of the results of the two large nationwide sexuality surveys I mentioned above. Further, there is a reason that the HIV-negative gays may not do as well as he thinks--they are not married to a woman. Considerable evidence suggests that being married to a woman is associated with more health benefits than income level or educational attainment. And, if Rosenberg were to read our paper instead of "shooting from the hip" in an interview, I suspect he would modify his tune.

We also have analyzed the age distributions of the married and homosexually partnered in Denmark, Sweden, and Norway. These age distributions offer another test of our thesis. In each of these countries the median age of the married is about 50 and about a quarter of the married are old. But among the legally partnered gays and lesbians the median age is about 40 and less than 5 percent are aged 65 or above.

Another index of homosexual life span comes from Colorado. Over the past three years the state Department of Health has recorded 7,638 HIV tests obtained by gays. Mostly these tests are given to the "worried well," since less than 7 percent of the tests indicate infection with HIV. Unlike the two national sex surveys, this is not a random sample but, like the sex surveys, there is a numeric "bulge" in the late 20s and early 30s. 75 percent of all gays who got HIV tests were aged 39 or less and only 1 percent of gays getting tested were old. The same kind of age distribution, with the same kind of bulge, was recorded for the 6,737 IV-drug abusers, and less than 1 percent of them were old.


The similarity of the gay and IV-drug-user age distributions is important. Our review of the forensic and sociologic literature has led us to believe that the median age of death for IV-drug abusers (IVD) is probably somewhere between the late 30s and early 40s (again assuming that they don't "drop out"). A few relatively brief longitudinal studies of IVD further confirm our estimate. Interestingly, both lifestyles so intimately associated with modernity, the homosexual and IVD, appear to result in a sharply reduced life span. Between the two, IVD appears to be the more costly in "years of life lost" but only by a few years. So finding considerable age-distribution similarities in these two large sets of gays and IVD also lends support to our initial thesis.

All this leads to some questions for Olson. Why did two nationwide sex surveys, self-advertised as "definitive," record so few older homosexuals? Why do legally partnered homosexuals trace such a young age distribution? Why are so few older gays taking HIV tests?

A plausible answer to these questions seems to be "because not many homosexuals become old." The same logic--that what they do causes the difference--is regularly invoked to explain why the age distribution of smokers is a few years younger than the age distribution of nonsmokers.

There are problems with our--as with any--empirical study. There may be mechanisms other than death to account for the lack of older gays. But given the multitude of self-inflicted medical (e.g., hepatitis, anal cancer, AIDS), psychiatric (e.g., suicide, homicide), and other social problems (e.g., 12 percent of gays with AIDS have IV-drug use as an additional risk factor), it is neither difficult nor implausible to imagine that many of them are simply "missing in action."


--Dr. Paul CameronFamily Research Institute

Walter Olson replies:

Yes, it's Cameron against a world of skeptics. It seems the whole health-science establishment--professors of medicine, AIDS epidemiologists, practicing doctors with gay clienteles--has up to now failed to notice that even HIV-negative gay men die on average by their mid-40s, nearly as young as those who do have the virus. Indeed, as Cameron's Omega paper concedes, the mainstream, peer-reviewed scientific literature contains many assertions to the opposite effect: that homosexuality is not in itself associated with a shortened life span, let alone one shortened by decades. University of California, Davis psychologist Greg Herek has just published a very severely worded 30-page critique of Cameron's work on homosexuality, as part of an essay volume in the (peer-reviewed) Sage series. One side or the other of this controversy--the mainstream literature and Herek's paper, or Cameron's paper--is hopelessly wrong, yet has slipped past the not especially fine-toothed comb of peer review.

Which should shock no one: It's no secret that lots of marginal output can find a berth in one or another of the superabundant supply of academic journals these days. To judge from its table of contents, Omega: Journal of Death and Dying caters to those who publish in such areas as bereavement counseling, cultural studies of death, and other matters far removed from Centers for Disease Control and Prevention-style epidemiology. Along with its substantive critique, Herek's paper proffers "a citation count analysis which shows that Cameron's work has been virtually ignored by social and behavioral scientists." It has, however, circulated widely among religious-right groups, which appear to view Cameron's willingness (elsewhere) to condemn homosexuality in very strong terms as a proxy for the scientific rigor he brings to the task.

Cameron now invites attention to the second, literature-survey portion of his article. It's fully as risible as his obituary survey. In this section, he adduces numerous bits of data regarding the age distributions indicated in various earlier surveys on homosexuality. These snippets not only are at hopeless variance with each other but repeatedly, and fatally, undercut his own thesis. Start with the 1858 data of whose antiquity he's evidently proud: a report by G. Tardieu on the age distribution of males imprisoned for sodomy in France. You'd expect such numbers to be skewed toward younger men, who 1) get in more legal trouble, 2) engage in more impulsive sex, and 3) are more likely to be in the military. Yet of subjects aged 18 and higher, the median age was "almost 40"--a figure inconsistent with the thesis that the average age of death is only a few years higher than that.

Again and again Cameron seems unaware that figures showing average or median ages for gay populations in, say, the high 30s, refute rather than confirm his notion that average-age-of-death is only a few years higher than that. Thus he triumphantly points to a subscriber survey for the gay magazine the Advocate, which reported an average subscriber age of 38. (To clarify the implausibility here, imagine a "typical" reader: subscribe at 18, die at 43, and then--to get the average subscriber age up to 38--go on subscribing for 15 years after death.)

Modern Western survey data find that younger respondents self-identify as gay in greater numbers than those of higher ages, not only because AIDS has carried off some older gays but also in part, clearly, because of a generational revolution: Those who've grown up in an age of reduced stigma are more willing to self-identify in surveys. Thus the British survey Cameron cites in his letter, which evidently was more pointedly worded than the American, is consistent with younger respondents being more sexually active or more willing to level with surveyors; it is quite inconsistent with his theories on average ages at death.

It all recalls the study a few years back on supposed high mortality among left-handers: Researchers found that between one survey and another the number of people reporting themselves as left-handed dropped sharply, and decided that the reason was not (boringly) a defect of the survey mechanism but (excitingly) that lefties had been dying like mayflies. The study has now mercifully been forgotten, but we'd still be debating it if Bill Bennett had decided to go on television to crusade against left-handedness.

As one who's put in nearly two decades working with such outfits as the Manhattan Institute and the American Enterprise Institute, and who's published in the last two years in such outlets as the National Review; the editorial pages of the Wall Street Journal,the New York Post, and Investor's Business Daily; the American Spectator; and Commentary, I must thank Cameron for the other bit of unintended mirth provided by his letter, namely his description of me as "on the left." It will go directly into my souvenir file.

I Was an Essayist for NPR

It is ironic that in "I Was a Teen-Ager for the New York Times," his extraordinarily nasty article about Joyce Maynard, Alex Beam expends more than a thousand words on a woman whose life he deems uninteresting. It seems to me that the envy he admits to in the beginning of his piece is still potent. He berates Maynard for writing about her life. Has the conventional wisdom changed? Isn't a writer supposed to write about what he or she knows? Or is it better to write a vicious article about someone else's life?

Alex Beam also neglects to mention Maynard's fine work as an essayist for National Public Radio's All Things Considered. I've had the pleasure of being her editor here for the last five years. She is a gifted writer, whose work is filled with insight and humor. She is also one of the most delightful people I've ever worked with.

--Margaret Low Smithsenior producer, All Things Considered

Thirteen Ways of Looking at Wallace Stevens

In "The Emperor of Ice-Cream," his review of the new Library of America volume of Wallace Stevens' poetry, Alex Ross writes, "Academic interpreters have failed to meet the challenge posed by Helen Vendler ... --to give up the search for intellectual subject matter and to treat Stevens as 'pure sound.' " This challenge (and I wonder if Vendler really made it) is all but impossible, and Ross fails it himself.

He discusses the sound of just one line, the first line of "The Idea of Order at Key West." He notes that it is iambic pentameter and that its vowels seem to have been carefully chosen--which does nothing to distinguish it from Spenser or Yeats. For the rest, he ignores sound. (He does remark that Stevens favors monosyllables, but this is not a matter of "pure" sound, since words run together in connected speech.) Almost all Ross' comments are based on meaning, as when he finds that the image in this poem becomes vague, or when he answers the rhetorical questions in "The Man on the Dump."

But, really, why would one want to treat Stevens as "pure sound"? His "intellectual subject matter" may not be marshaled as finished argument, but it is hard to miss. ("Key West," for instance, includes a comparison of art and the world.) I suspect this "challenge" is just an excuse for Ross to put down unnamed "academic interpreters," using Vendler as a handy authority before flunking her, too.

--Vance MaverickSan Francisco

If we're going to have a debate, let's at least get the facts right. For starters, in "Speed Trap," Paul Krugman misspelled my name. I'll get over it. More important, he is just as careless in depicting the arguments of those of us who think the economy is now capable of decent growth without reigniting inflation.

My essay ("The New Economy: What it Really Means") in Business Week was much more balanced and modest in its claims than Krugman suggests. I invite Slate readers to reread my piece rather than accept Krugman's caricature of it, then decide for themselves. Business Week has generally been right about the performance of the economy in recent years--and right in urging the Federal Reserve not to raise interest rates. If the Fed had raised rates in order to hold growth down to 2.5 percent over the last 18 months, as Krugman seems to prefer, we would have given up $250 billion in economic output. The unemployment rate would have been at least a half-percentage point higher, putting more than 750,000 people out of work.

1997 was a wonderful year--strong growth, declining inflation, declining unemployment, rising productivity, and rising real wages. Happy New Year, Professor Krugman.


"The New Business Cycle" (Business Week cover story, March 31, 1997)

"How Long Can This Last?" (Business Week cover story, May 19, 1997)

"Alan Greenspan's Brave New World" (Business Week cover story, July 14, 1997)

--Stephen B. Shepard

editor in chief, Business Week

Paul Krugman replies:

First of all, my extreme apologies for the spelling mistake (The error has been corrected.).

As for substance: We now see just how helpful an Internet magazine can be. By all means read the original Shepard essay in Business Week (notice that I did include a link to that essay in my original piece), as well as the earlier articles on the site, and ask yourself whether they did or did not appear to claim that our favorable measured growth and inflation performance can be explained by unmeasured productivity growth. If they did claim or even suggest that, then Business Week has indeed been saying something very silly. Oh, and by the way: Reread my own piece, and ask whether I actually say that growth over the past 18 months should have been held down to 2.5 percent, as Shepard claims I "seem to prefer."

What I do know is that a number of Business Week readers have contacted me since reading my piece and told me that the explanation of the speedometer fallacy was new to them--that they were under the impression, from reading BW, among other sources, that a hidden productivity boom was not only a possible explanation of low measured inflation but indeed the leading candidate.

I guess my question is whether Mr. Shepard is willing, even now, to concede that it is a fallacy to invoke unmeasured productivity as the source of our good numbers. (And I don't mean a judicious acknowledgment that there may be something to what I say--we are talking 2 + 2 = 4 here, with no wiggle room whatsoever. Extremism in the defense of arithmetic is no vice.) If he is, then I am at least prepared to consider the possibility that I (and everyone else I know) have been misinterpreting all those articles. But I would like to have the point acknowledged.

Address your e-mail to the editors to letters@slate.com. Please include your address and daytime phone number (for confirmation only).