A Misleading Poll on Human Genetic Technologies

If you ask the wrong question, you will certainly get the wrong answer. Likewise, if your questions are confusing or misleading, you may get answers that don’t make sense. In the worst of all worlds, the answers, right or wrong, can be misinterpreted to say something they clearly don’t.

The Pew survey of global attitudes to biotechnology research published on December 10 asks the wrong questions in a confusing way. This is significant, in itself and even more because all the discussion about the future legal status of heritable human gene editing includes the assumption that a properly informed public will weigh in on the issue. Confusion about what people will accept or reject serves no one in the long run.

From October 2019 to March 2020, researchers from the Pew Research Center conducted surveys in 20 countries, most by telephone, some by using face-to-face interviews, for the International Science Survey 2019–2020. The results are analyzed and published by Pew in at least two different reports. One selection was published in September as “Science and Scientists Held in High Esteem Across Global Publics.” The latest takes as its general topic Science and Religion. The first half, however, is not specifically about religion but rather whether the individual surveyed considers particular applications to be appropriate; the second half does specifically address religious beliefs. The report is titled:

Biotechnology Research Viewed With Caution Globally, but Most Support Gene Editing for Babies To Treat Disease

The problems start right there. The intention of this part of the survey is not to discuss gene therapy to treat sick babies, but rather gene editing of embryos that will become future babies, if all goes well. Which it very well may not. At no point does the report even clarify the distinction between somatic (affecting living patients) and germline (heritable) applications. The assumption throughout appears to be that the discussion concerns germline gene editing, and the failure to even mention this is an important omission. If the distinction between healing actually existing babies and creating hypothetical genetically modified babies had been made clear, it probably would have made a substantial difference to the survey results.

The report does mention animal cloning, which is (as usual) unpopular, but the gene editing discussion is entirely focused on humans. Other than that omission, the summary is not essentially inaccurate:

Global publics take a cautious stance toward scientific research on gene editing, according to an international survey from Pew Research Center. Yet most adult publics (people ages 18 and older) draw distinctions when it comes to specific applications of human gene editing, including showing wide support for therapeutic uses.

Indeed, the report quite rightly stresses the strong opposition to changing a baby’s genetic characteristics to make the baby more intelligent, which 82% of those surveyed consider to be “a misuse of technology.” It does, however, find support for “scientific research on new technologies to help women get pregnant.” In general, younger people are more supportive of research into gene editing, as are those not affiliated with a religion. There are some interesting differences in national responses, with people in India being the most supportive of genetic technologies.

The report goes astray over the question of preventing the transmission of serious genetic conditions to the next generation. This is a sympathetic endeavor, but gene editing to accomplish it is completely unnecessary. Yet the report includes no mention of technologies that can and do provide parents with genetically related children who are not affected by the condition they are trying to avoid.

Preimplantation genetic diagnosis (PGD) enables the selection of embryos rather than the editing of them. This does raise ethical and practical questions, which for decades have been pointedly raised by disability rights and justice advocates, and discussed in academic, professional and governmental circles. But the prospect of embryo editing – modifying the genes and traits of future children and generations – raises the concerns about genetic selection to a whole new level.

Asking whether awareness of alternatives to heritable genome editing would affect parental and public support for it seems essential to an accurate portrayal of opinion. Not even mentioning the concept is at least a missed opportunity, and at worst a way to stack the decks.

Rather than making sure that respondents understood the context of the issue on which they were being asked to weigh in, the survey asked a short series of questions that culminates with this one (Q23c):

Would you say it is appropriate or misusing technology to change a baby’s genetic characteristics ____? c. to TREAT a serious disease or condition the baby would have at birth

Read carefully, the phrase “would have” makes it quite clear that the technology is being applied before birth. Therefore, not to a baby. Perhaps in utero, to a fetus. Indeed, the question itself is self-contradictory: If you are changing a baby’s characteristics, then it is too late to apply treatment before birth. As noted, the report does not even discuss the distinction between somatic and germline gene editing. (There is one mention of “germline” in the context of the widespread condemnation of He Jiankui, in part because of unknown health complications.) None of the questions address this.

Pew has conducted surveys about Americans’ attitudes on these topics before, in 2018 and 2016. Those earlier surveys did at least mention embryos, and in 2016 Pew did ask some questions that edge in a slightly more interesting direction. That year’s survey could also be critiqued, for similar reasons, but it did include some implied references to heritability and a question about “testing on human embryos in order to develop these techniques.” Some general sense of people’s attitudes can be drawn from the 2016 survey. The 2020 edition, however, simply fails in that task.

Finally, there is a glaring contradiction in the responses to two different questions. Pew’s summary states that:

A 20-public median of 63% say scientific research on gene editing is a misuse – rather than an appropriate use – of technology.

This summarizes Q12a, which specifies “gene editing to change people’s genetic characteristics.” However, the responses to Q23c (quoted above) would seem to contradict that:

A 20-public median of 69% say it is an appropriate use of technology to change a baby’s genetic characteristics to treat a serious disease or condition the baby would have at birth.

[As noted above, the first use of “baby” here means “embryo” in practice.]

Comparing the two, a significant number of respondents, perhaps a majority, say it is appropriate to misuse technology on embryos. Somehow, I don’t think this is what anyone intended.

Perhaps the pollsters should reconsider their methodology.

Pew is one of the few organizations with the resources to conduct such a poll, and accurate information is essential to a full discussion of the debate about heritable human genome editing. CGS has collected data going back to 1986, and this survey does not contradict the general pattern, but its limitations mean that it really does not advance our understanding. So maybe it’s just as well that, almost a week after publication, no major newspaper has covered it.