"Consumers Warm to Gene Editing": Where Is the Evidence?

Farmers Weekly ran a story on 28 September under the headline Consumers warm to gene editing in crops, but animals a harder sell. The first line says most consumers accept gene editing in plants. The source is one speaker on one conference panel. The story names no survey, no sample and no country.

The speaker is Stuart Smyth, a professor of agricultural economics at the University of Saskatchewan. The story says he was on a panel at the World Agri-Tech conference in London.

I went looking for the evidence. There are four problems. I'll take them in order: what the story offers, what the public actually knows, how the surveys asked, and who paid.

1. What the story offers as evidence

The evidence offered is area planted: gene-edited plants across South, Central and North America, Japan and Australia. That tells you what farmers plant and what regulators allow. It tells you nothing about what shoppers think. Hectares are a fact about tractors.

The story also treats gene editing and the older genetic modification as one thing. A 2022 paper Smyth co-wrote says the first genetically modified crops (canola, corn, cotton and soy) reached farms in the mid-1990s. As of 2022, it described gene-edited apples and potatoes as products only starting to appear in Canadian shops. So it is reasonable to suspect most of those hectares are the older technology. That is my reading. The story doesn't say.

2. Surveys keep finding people don't know what gene editing is

Start with the terms. As the papers I read define them, gene editing changes an organism's own DNA, and genetic modification (GM) brings in DNA from another species.

Now the public. In Smyth's 2022 survey of 497 Canadians, 72% rated their own knowledge of gene editing as poor or very poor. A 2026 US experiment found 91% had heard at least a little about gene editing in agriculture. Before watching anything, they averaged about one correct answer out of three on knowledge questions. That average is my reading of the paper's table.

New Zealand looks the same. Plant and Food Research's own write-up, from July 2023, says its team found most consumers don't differentiate between gene-edited and GM technologies. In the RNZ report on that work, programme leader Dr Denise Conroy says focus group members felt they didn't understand gene editing or genetic modification because, as they put it, New Zealand hasn't had a public debate on the subject for about 20 years.

So "consumers accept gene editing" rests on people who, by their own account, mostly can't say what it is. You can't meaningfully accept something you can't describe.

3. The studies I could read explained it first, then asked

In a courtroom, "leading the witness" means asking a question that suggests the answer. A survey can do something similar. If you explain a technology first and then ask what people think, you measure their reaction to the explanation. That is a fair thing to test. It is not the same as what people think today.

Here is what each study did.

Smyth's 2022 survey. Respondents were given definitions of gene editing and GM before they were asked their views. The result that gene editing beat GM (29% positive against 16%, with half neutral on gene editing) came after that. The authors themselves caution that a campaign built only on educating people could lead consumers into confirmation bias.

The New Zealand milk study. Before the choice experiment, respondents were told the milk came from cattle developed with gene editing to improve traits such as added value and environmental sustainability, and were given a table explaining each attribute. The product was called "climate-smart milk" instead of gene-edited. The authors write that this framing may have raised preference. The gene-edited versions were priced in the same range as standard milk on an assumption of cost savings, which the authors call an exploratory scenario. Standard milk was still the most preferred. At equal prices the simulation gave it 52%, gene-edited milk 27% and organic 21%. The data was collected around January 2023 and the paper was published in June 2026.

The US experiment. It was built to test which messages shift opinion. The number before any message was, by my addition of its table, 48% calling gene editing in agriculture somewhat or extremely safe, 30% unsafe and 22% neutral. Fewer than half called it safe. After a 30-second video, opinion rose by about half a point on a five-point scale. There was no comparison group without the video, and the videos used the word "bioengineered".

A 2021 Japanese experiment of 4,514 adults, run in March 2018, found the picture shown first changed attitudes to gene-edited livestock. It also found gene-edited vegetables seen as more beneficial than livestock, so Smyth's split between plants and animals has some support. The story just doesn't cite it.

Plant and Food Research says naming and framing can change preference significantly. Its write-up doesn't say how its 43% figure was asked. That figure is the share of New Zealand and Australian survey respondents willing to buy produce from gene-edited technology. It was the most rejected technology tested. Vertical farming scored 62%. Neither the write-up nor RNZ says how many people answered or who paid.

By my count, all four studies I could read in full explained the technology to people before measuring opinion. The fifth doesn't say. The number closest to what people think today is the one taken before any message, and it is the US baseline.

4. Who paid: a guide to the conflict of interest

A conflict of interest is a financial tie that could reasonably be seen as pulling on someone's judgement. It doesn't prove the judgement was pulled. The declaration in Smyth's paper uses the standard test: any commercial or financial relationship that could be construed as a potential conflict.

Smyth. His university funding page lists current funders, as of 2025, including Bayer CropScience, CropLife Canada and Syngenta, plus grain grower groups. It describes Bayer CropScience as a crop science company in seeds and crop protection, and Syngenta as an agriculture company. Monsanto is listed among the partners of an earlier chair, 2014 to 2019.

The page describes CropLife Canada as the trade association of makers of plant science technologies, including plant biotechnology. It lists among that group's priorities raising public awareness, understanding and acceptance of the benefits and safety of those technologies. The 2022 paper Smyth co-wrote studies public acceptance of gene editing. The paper lists a different grant as its funding, so I can't tell whether any of the CropLife money paid for it.

The paper's declaration says the authors have no relationships that could be construed as a potential conflict. The Farmers Weekly story mentions no funding for the speaker. It does list the funders of its own series, including Beef + Lamb NZ and the Meat Industry Association.

The milk study. The milk study's acknowledgements say it was funded by AgResearch, CRV Ltd, Livestock Improvement Corporation and the Ministry of Business, Innovation and Employment. LIC describes itself as a farmer-owned co-operative and lists bull teams and breeding genetics among its products. CRV's New Zealand site sells bulls and genetic services. The paper thanks its project steering group for feedback on the survey design and doesn't say who is on it.

The paper calls itself a preliminary check on whether investment in gene-edited cattle would find a market. It says consumer acceptance data can guide that investment.

The third author, Götz Laible, is the last-listed author of a 2018 AgResearch paper reporting gene-edited calves engineered to stop producing a major cow's milk allergen. AgResearch's own page lists him as a Principal Scientist there and an honorary Associate Professor at the University of Auckland, and says he joined in 1997 to develop technologies for the genetic improvement of livestock. The milk study cites the 2018 paper as support for its allergy-free version. The authors' own summary, republished by Yahoo News Australia, lists Laible as receiving ministry funding with additional support from CRV and LIC. The paper declares no conflicts of interest.

What this doesn't show. Nothing I read shows a funder shaping a result. Industry funding of research is common, and the milk study still found standard milk preferred. The point is narrower. In my view, these ties meet the "could be construed as" test, and readers were entitled to be told. Farmers Weekly didn't tell them.

5. Figures with no source

The story attributes food waste of about NZ$2100 per household to unnamed researchers. It says edited plants could sequester 10% more carbon, and that parts of Asia have eased controls. It doesn't say where any of that comes from.

How to read the next one

Four questions do most of the work. Who paid, and do they sell the thing being studied? What were people told before they answered? What was the product called? Was the number taken before the explanation or after it?

Smyth's message, per the story, is "Listen to the silent majority." A silent majority is, by definition, one nobody has heard from. In the survey he co-wrote, most people said they knew little about gene editing. That was the finding. The headline says they're warming to it.

Before you go

If this kind of check is useful, sign up free and get the next one by email. If you'd like to help keep it independent, you can leave a one-off tip.