Relative danger

Natural History, Sept, 1997 by Colin Tudge

Tell any group of people that stepparents are about one hundred times more likely to fatally abuse their children than are "true" parents and you can't expect an easy audience -- especially if you suggest that the key factor could just be the lack of a genetic relationship. So the questions rained down thick and fast on Canadian psychologists Martin Daly and Margo Wilson when they floated these conclusions (gleaned from their twenty years of research) at a Darwin Seminar at the London School of Economics. The questions were as varied as the audience of scientists, historians, undergraduates, and interested passersby that these provocative seminars invariably attract.

"How can you know such a thing?" people asked. "What makes you think that genetic relationships play a part, as opposed to a hundred other possible confounding variables, such as the poverty of the parents and the duration of contact between parent and child?" "What about adopted children?" "What use is such information, anyway?" "Aren't you just stirring things up for no good reason?"

Daly and Wilson, from McMaster University, Hamilton, Ontario, did not arrive at their conclusions casually. In the 1970s, about ten years after the "battered child syndrome" had been officially recognized, they set out to see if children were more likely to be abused by stepparents than by biological parents. In addition, they wanted to explore a specific Darwinian hypothesis. There had been new developments in Charles Darwin's idea that human behavior, as well as our physical appearance, has been shaped to a significant extent by natural selection. In particular, in the 1960s, William D. Hamilton, now professor of zoology at Oxford University, had put a new construction on the phenomenon of altruism -- the process by which individuals apparently sacrifice their own self-interest, and even their lives, to benefit others.

Hamilton showed that natural selection could and would favor genes that promoted altruistic behavior, provided that the individuals who benefited from the altruism had a high chance of containing the same genes. In fact, such altruistic behavior is not altruistic at all, in the sense that moral philosophers use the term, but is entirely selfish. The gene that promotes the apparent self-sacrifice is simply promoting its own replication, by enhancing the survival of copies of itself -- albeit copies contained in other individuals. Thus, said Hamilton, we might expect individuals to compromise themselves if doing so benefited their own kin, who would indeed contain copies of the gene that promoted the self-sacrificial behavior. The final mist is that parental care, and the self-sacrifice that goes with it, are merely special examples of the altruism that any organism might be expected to show toward its own kin.

Although stepparents in some societies are related to the children, Daly and Wilson reasoned that stepparents are not generally kin to their stepchildren, at least not in most Western societies. Therefore, we might expect that they would show no predilection to sacrifice themselves (in large or even small ways) on a stepchild's behalf. They sought to find out if this hypothesis, based on Hamilton's extension of Darwin's ideas, was true.

Wilson says, "We were astonished to find that it was not easy even to begin to explore this hypothesis. Official statistics from the United States didn't reveal whether parents who abused children were step or biological. It just didn't occur to criminologists that the nature of the relationship was important, so they generally didn't bother to record it." She and Daly had to took beyond the official statistics, to the raw data of case histories. By 1980 they had demonstrated that children under three years of age are at least seven times more likely to be abused by stepparents than by biological parents.

Daly and Wilson believed, however, that statistics for child abuse in general might be biased by underreporting or incomplete reporting. After all, parents don't want to admit that they have beaten their children; there are plenty of ways to conceal abuse or to explain away injuries. To gain a truer picture, the researchers decided to focus specifically on a form of abuse that is exceedingly difficult to cover up: homicide. Once again -- even more astonishingly -- most official statistics, including the FBI's Supplementary Homicide Reports, the U.S. national archive, did not differentiate between killings by stepparents or by biological parents. But Statistics Canada from 1974 to 1990 did contain relevant data.

Its figures showed that children under the age of two were at least one hundred times more likely to be killed by stepparents -- particularly stepfathers -- than by biological parents. "Of course," Daly stresses, "most stepparents take to the task extremely well, and generally make loving substitute parents. The incidence of abuse is low." Nonetheless, for stepparents the homicide rate comes out at about 600 per million parent-child groups living together, compared with just a handful for biological parents. Further examination of records in the United States and Britain revealed an increased risk for children with stepparents. Moreover, a closer look at the case histories reveals that while biological parents who kill their children are often severely depressed and, Daly and Wilson reported, "may even construe murder-suicide as a humane act of rescue," stepparents who are homicidal "are rarely suicidal and typically manifest their antipathy to their victims in the relative brutality of their lethal acts."