Dr. Andrea Perrella and I have a new paper in the Canadian Journal of Public Health that reveals some interesting insights into how people process information and weigh the risks and benefits of water fluoridation. In particular, we found, through an experimental survey instrument, that people’s support for water fluoridation can be lowered, but not raised.
In our survey experiment, we found that when people were told that Maude Barlowe and the Council of Canadians oppose municipal fluoridation (this is a source cue), respondents were significantly less likely to agree that municipal fluoridation reduces cavities (We were pretty surprised at this but here is a resolution opposing fluoridation and here is a link to Maude Barlowe reiterating the same position).
Why would this be? After all fluoridation is a remarkably effective and equitable public health intervention. Fluoridating the municipal water supply is a great way to reduce cavities in a universal and cost effective way.
There are two reasons for our result. First, people’s expressed attitudes in opinion surveys are not reflections of stable, independent attitudes that can be naively measured. Rather, they are expressions of a mix of conflicting attitudes, thoughts and feelings that people draw on to formulate a response to a public opinion survey. Second, people’s perception of risks is highly non-rational (see here). Ask people if crime is worse now than it was when they were younger and you will get a lot of people saying “yes,” despite the fact that this is almost certainly not the case (see here).
So, when we ask people whether they think fluoride reduces cavities, responses provided are a mix of conflicting durable and fleeting thoughts and feelings. Our study shows that if you provide people with one extra piece of information, such as, for example, that a well-known and largely well-regarded person and her organization (i.e., Maude Barlowe and the Council of Canadians) are opposed to it, and then subsequently ask whether they think it is factually true that fluoridation reduces cavities, even if that prior information is completely tangential to the scientific veracity of the claim or not, then you will drive down support for the factual claim.
In our case, we turned people’s responses to the statement into a scale from 0 to 1, 0 being strongly disagree 1 being strongly agree. Within the control group, which comprised individuals who did not receive a prompt, the average score was 0.636, but the average response in the Maude Barlowe/Council of Canadians experimental group was 0.565, a full 11% lower. Interestingly, we had a second experimental group of people who were told only that Health Canada and the World Health Canada supported municipal fluoridation. Interestingly, there was no increase in support for the belief that fluoride reduces cavities. In short, people basically express support for the statement that fluoride reduces cavities, but you can't increase that support, but you most certainly can decrease that support.
Now, it’s important to note, that even in the Barlowe/Council of Canadians group, the average response was above 0.5, so there was still majority support for the statement that fluoridation reduces cavities. That’s good news. Moreover, we followed the question about the benefits from municipal fluoridation with an item asking people whether they agreed or disagreed that fluoride in the municipal water supply can cause bone disease or skeletal problems. At very high doses, fluoride can, but this is far above the level at which Canadians are exposed to. On this question, there was no difference between the experimental and the control group, so it's possible that the impact of the source cue that was provided was a very fleeting one that really only lasted for one survey question.
Nevertheless, these findings should give municipal decision-makers cause for concern when they are presented with demands for plebiscites on water fluoridation. If you have a plebiscite campaign where the news media are careful to only publish information that is justified by existing scientific evidence and where there is a vocal, organized and competent campaign to support fluoridation, then we expect voters might get enough cues to make good decisions about fluoridation. But if, through whatever contingencies, local news media provide tangential information (like mentioning the fact the Council of Canadians opposes fluoridation) or coverage to groups whose claims are not scientifically validated, and if fluoridation opponents organize well and commit resources to the campaign while the pro-fluoridation side fails to do this, then our evidence suggests voters could easily be led astray and a remarkable public health intervention might be reversed. After all, all you need to win a plebiscite is 1 more vote than the other side.