I just read in an academic paper (Macintyre and Sooman, Lancet, 1991) that in modern populations, the cuckoldry rate -- i.e., the rate at which men are deceived into raising offspring that are genetically not their own -- is 10 to 15 percent. This would make genealogy and family reunions a moot point. What's the straight dope?
Curious, via e-mail
Cuckoldry: the word alone – conjuring as it does some Chaucerian vignette of a smiling wife and her paramour exchanging sly glances under the witless husband’s nose – well conveys the issues in play here. The notion of a man’s being duped this way has for millennia inspired a potent mix of anxiety (on the part of potential cuckolds) and snickering (on the part of everyone else), and thus any suggestion of its pervasiveness is bound to attract a lot of attention, as it’s apparently attracted yours.
Perhaps for this reason, some possibly iffy statistics on cuckoldry – or nonpaternity, as the killjoy experts usually call it – have wound up with an outsize place in discussion of the matter. In fact, that’s a major point of the article you mention. Sizing up the conventional wisdom on nonpaternity before the advent of widespread DNA testing, the authors find that though nonpaternity rates of 10 percent and higher have routinely been cited in studies and textbooks, these numbers turn out to have little solid data behind them. Among the estimated rates they found:
- More than 30 percent – obtained from a researcher’s remarks at a 1972 symposium on medical ethics, referring to a study (apparently conducted decades earlier in a single English town) that was never completed, much less published.
- 20 to 30 percent – from another aging and unpublished UK study; and
- 7 to 14 percent – from a 1990 study that relied (as later researchers would point out) not on any biology-based testing but on self-reporting by readers of a British women’s magazine on the frequency and timing of their off-the-books intercourse.
It’s hard to believe such rickety numbers would’ve featured in the discussion at all if their implications hadn’t been so juicy and there’d been more reliable numbers to focus on instead.
Roughly 20 years into the DNA-test era, better paternity data are available and scholars have devised better ways to break them down. For a 2006 survey anthropologist Kermyt Anderson took 67 studies that estimated nonpaternity rates and sorted them according to “paternity confidence”: the high-confidence group included, e.g., nonrandom genetics studies of parents and children, which (Anderson reasoned) families would be unlikely to volunteer for if someone suspected the results might show that dad wasn’t the kids’ father; in the low-confidence group were straight-up paternity-dispute test data; and a third category contained studies from which one couldn’t conclude anything about fathers’ confidence.
Seen this way, the numbers yield a pretty convincing pattern. The median nonpaternity rate for the high-confidence group was a not-too-scandalous 1.7 percent, whereas the low-confidence group showed an unsurprisingly high rate of 29.8 percent – about what one might gather from watching a few weeks of Maury Povich. If you combine the first group with the can’t-conclude group, which showed a rate of 16.7 percent, you get a rate around 3.3 percent, or a ninth of the low-confidence rate. While Anderson cautions that there’s currently no way to figure out what percentage of total births are low- or high-confidence, and thus what a societywide nonpaternity rate might be, he does use figures from a paternity confidence study he conducted in Albuquerque to guess that the rate for that city as a whole would be under 4 percent.
Such a figure squares a lot better with other recent large-scale surveys than the 10 percent rate one hears. In a 2005 paper Australian sociologist Michael Gilding reads available evidence as suggesting a nonpaternity rate for Western countries of between 1 and 3 percent; another comprehensive study of international data agrees we can’t yet draw any conclusions about across-the-board rates but says that minus paternity-dispute cases the overall rate looks to be about 3.7 percent.
While papers focusing on specific population segments may be of limited use due to sample-size issues, some fascinating small-scale research also supports the macro findings. For a 2000 study Oxford scientists collected DNA from 48 men with the last name Sykes living in a particular section of northern England. On genotyping the subjects’ Y chromosomes, the researchers found that (a) unexpectedly, there seemed to have been a single ur-Sykes with whom 44 percent of the living Sykeses shared a unique string of genetic info, and (b) over 700 years the Sykes nonpaternity rate had been about 1.3 percent per generation. Now, if one Sykes’s wife got together with another Sykes on the side, any resulting nonpaternity wouldn’t show up here. But assuming that women who married Sykeses were neither atypically unadventurous nor surprisingly prone to Sykes-swapping, this too suggests that cuckoldry isn’t nearly the undying scourge it’s been made out to be.
Send questions to Cecil via firstname.lastname@example.org.