When Faunalytics carried out research on veg*n recidivism back in 2014, no one could have predicted the way one headline stat – that 84 percent of vegetarians and vegans abandon their diet – would dominate the agenda and stick in the minds of those working for (and against) animal protection. But that startling figure, which has shaped the strategies of major advocacy organizations and made its way into popular consciousness, did not tell the whole story.
The research, from a well-respected analytics organization founded in 2000 that provides robust research to animal advocacy groups, went viral. Authors such as Hal Herzog praised the research, claiming it showed that 'animal activists should emphasize reduction, not elimination, of eating meat'.
The figure continues to be used to support a wide range of vegan, reducetarian and incremental approaches to diet today; for example, the new UK campaign Middle Ground explicitly employs the figure to support its focus on those who care about animals but are unwilling to wholly give up animal products.
The outcome of that 2014 study continues to act as a standard for Faunalytics' approach to all of its research.
"We consistently recommend against absolutism in advocacy," says Jo Anderson, Research Director for Faunalytics, "as it promotes a sense that an aspiring vegan has already failed if they slip once or aren't 'quite there', rather than a sense of pride and continued motivation at a 99 percent accomplishment."
As Anderson points out, if the research suggests that one in eight Americans (the study was conducted in the U.S on American citizens) have trialled the diets, but only two percent of Americans currently identify as vegetarian or vegan, then much more attention needs to be given to adherence to dietary veg*nism rather than adoption of the diets.
And, as Anderson continues, an 'absolutism' or 'abolitionist' approach to advocacy is often a part of why those who give the diet a try slip back into eating animal products, whether dairy or meat. (I found this to be true in my qualitative research on men and veganism: if it’s a 'pass or fail' task, then one 'mistake' can lead to the whole attempt being given up.)?
Can the research be trusted?
There were many criticisms of the research at the time – and of the 84 percent figure – that give reason to question if we can use the data to support approaches to vegan outreach today.
Is veganism only a diet? (The study was limited to dietary veg*ns only.) Can you call someone who has tried the vegan diet once for less than three months a 'vegan'? If so, do the vast majority of these 'vegans' and 'vegetarians' really give up their diet, or do they go back and forth? And if they do, what does that imply about how vegan activists and organisations need to reshape their efforts to be effective in convincing people to change their lifestyles?
"Before we can address those questions with research," warns Anderson, "we also need to answer what we consider 'effective'. There are different ways to define effectiveness, and maintaining a 'pure' vegan or vegetarian diet may be the gold standard, but another key takeaway from the 2014 study is the importance of supporting efforts to reduce, semi-vegetarianism, flexitarianism, and other diets that reduce overall levels of animal suffering."
For example, one of the outcomes of the research was that those veg*ns who did 'abandon' their 'pure' vegan or vegetarian diets still had a diet reduced in animal products, especially meat, afterwards; and that 37 percent of those who had tried the diets at least once would be willing to do so again, with health the majority reason.
One of the other critical findings was that most of those who gave up their diets had little to no support from a veg*n community or network. That’s a reason, perhaps, why the Challenge 22+ model (with mentors) is proving effective.
However, those figures were not so widely reported, or remembered.?
Thinking about motivations
This argument against 'absolutism' in advocacy in favor of supporting a wide range of options and strategies, such as flexitarianism, Meatless Mondays, and even Veganuary, upsets some animal and vegan activists who consider it a watering down of the vegan message, and a betrayal of the animals suffering in the food production industries.
"The criticism of the research mainly seems to come from the abolitionist angle," says Tobias Leenaert, author of How to Create a Vegan World, "who don't seem to want to believe it or who wanted to see in it a confirmation of 'animal rights' only advocacy."
When the research was originally published, Leenaert wrote at length on the findings, arguing:
It's all good and well to say that people should be motivated by ethical reasons, but that doesn't mean they easily will be. It seems to me common sense – and is repeated in much of the change literature I come across – that it is more productive to formulate our message in a way that it connects to values that people already have (health, environment), instead of trying to get them adopt the values we would like them to have. ?
He continues to make the point today, telling me that: "People who get into veganism for health reasons can shift to animal reasons, right?"
His argument is that if vegan activists only use 'pure' or 'animal-only' motivations to encourage people to become vegan, then change will happen more slowly. Even though he agrees, and research suggests, that more people 'stick' with veganism if motivated by animal rights reasons, "but what if there would then be less of an inflow of people hearing those reasons, because you don't use those other (health, environment) arguments to get them?"
"I think we need to ask very specific questions about how we can help people maintain their diets," agrees Anderson, and a new planned study from Faunalytics will attempt to do just that. "Meantime, we have emphasized the how to other organizations in the years since the study, but there still hasn't been a lot of research on it."
Yet there remain flaws with what current research we have, which need exploring if we’re going to improve our advocacy for animals and grow more fully vegan societies. Do 84% of veg*ns really give up their diets, or, as some argue, is the figure likely to be much lower?
Let's look at some of the issues:?
#1: Vegetarians and vegans are not the same
Perhaps one of the major flaws in the research is the lack of distinction given to vegans and vegetarians in the reporting of findings.
Anyone who has made the switch from vegetarian to vegan knows that even if the weekly shopping basket is not quantitively that different, the qualitative difference can be stark and emphatic. Even Anderson from Faunalytics discovered this on her transition from vegetarianism to veganism (or 'vegan-ish') which she expresses eloquently in this blog.
As Anderson writes: "I rolled along happily for years as a vegetarian, secure in that identity - but definitely avoiding information about the egg and dairy industries."
Dairy remains a major blockage to many making the dietary switch, writes Dr. Neal Barnard in his book The Cheese Trap, despite the fact dairy production is seen by many as a more exploitative industry than beef or other meat products.
It is clear that vegetarian and vegan identities and practices are very different (something I wrote about in my book The Pig in Thin Air). For meaningful research to lead to meaningful advocacy strategies, these identifications and practices must be demarcated and separated in future research. Anderson would go further:
"In our own work, I intend to expand the definition of effectiveness," she says. "We will report the percentages of people who remain veg*n over time, but also those who contain to maintain a reduced-meat diet."?
#2: It conflates veganism with diet
"It’s really 'plant-based eaters', not 'vegans' who are 'lapsing' in the research," says Dr. Matthew Cole, vegan researcher and Lecturer in Sociology at the Open University. "If we contend (as I do) that veganism is an ethical commitment to oppose the exploitation of nonhuman animals, then veganism is likely to be far more robust than this research might imply."
In the Faunalytics study, the only reason given by a majority of those who had tried and then given up on vegetarian/vegan diets, for originally trialling the diet, was health (58 percent). This corroborates Cole’s argument: if a vegetarian or vegan diet is only adopted for health reasons, then it isn't veganism at all, but a plant-based activity.
Counter to Leenaert’s argument above, Cole suggests if the expected health benefits don’t materialise, then people look around for a different diet to follow – of those who then gave up, a third (34 percent) did so in the first three months, and another 53 percent before the year was up.
But Anderson argues this research is still valid, despite the lack of consideration as to whether you can call someone who has trialled a vegan diet for six weeks a 'vegan'.
"Whether you agree or not that people who were veg*n for only a short period of time should be included in a study of recidivism, every new veg*n has to make it past the first week, first month, first three months," she argues. "It's the most crucial time in behaviour transition so a very necessary research question, just a different one from what some people might have liked to see, if their goal was to find a type of veg*n for whom recidivism was low so it 'looks better'."
Indeed, such conflation of diets and identities happens on both sides. Positive campaigns such as Veganuary continue to equate veganism with diet only, with their starter kit focused exclusively on animal-based foods, but not clothing or entertainments, during January.
"But vegans, understood as those who oppose all animal exploitation, rather than just followers of a plant-based diet, would be much more resilient and less prone to 'lapsing'," argues Cole. "This is one reason why it remains important to contest the meaning of 'veganism' and to insist on its relationship to social justice."?
#3: It's not the biggest sample
And that’s the finding of another major study, where veg*ns self-identified and were found to have strong motivation for and stickiness to their chosen diets: the EPIC-Oxford Study.
The Faunalytics study from which the 84 percent figure comes was based, initially, on 11,429 North Americans. The follow-up qualitative work into the reasons for why people might give up their vegetarian or vegan diets was based on a subset of this: 1,387 respondents.
While these numbers are still reasonable for drawing results, they are far from the largest survey conducted exploring the 'staying power' of vegetarians and vegans. The EPIC-Europe Study explores the dietary choices and related health outcomes of over half a million (521,000) people across ten European countries.
In the UK there were two cohorts, in Norfolk and Oxford. The EPIC-Oxford Study grew out of the Oxford Vegetarian Study (which ran between 1980 and 1986) and ended up looking at the lifestyles of around 65,000 people. The study specifically over-emphasized vegetarian and vegan diets, which made up around 50 percent of all those involved.
Dr. Paul Appleby is a senior statistician at the Oxford Cancer Epidemiology Unit, and member of the Vegan Society's Research Advisory Committee, who worked on the EPIC-Oxford Study and continues to publish work based on its findings. The figures here suggest a very different story to those of the Faunalytics figure.
"Data from the EPIC-Oxford study shows that nearly three-quarters of the participants who were vegetarian or vegan at recruitment in the mid to late 1990s were still either vegetarian or vegan when they completed a follow-up questionnaire in 2010," Appleby told me in personal communication. That is, 73 percent of those who identified as vegetarian or vegan back in the 1990s were still following those dietary lifestyles over 20 years later.?
And the recidivism rate is slow. As Dr. Timothy Key, principle investigator on the EPIC-Oxford Study has shown, in their five year follow up after the study was first concluded, 85 percent of vegetarians and vegans were still following their chosen diet. So the recidivism rate after five years was not 84 percent, but the other way around, only 15 percent. And a further fifteen years later it was still not 84 percent, but only 27 percent.
Appleby is hoping to publish this data soon, after which the difference in recidivism rates between veganism and vegetarianism can be more clearly seen. But if you agree with Cole that veganism is a commitment to end the exploitation of animals, then it may be likely that much of that 27 percent would be made up of vegetarians, who were consuming animal products anyway, rather than the vegans.?
Does the EPIC study have flaws?
The participants in the EPIC-Oxford Study may have particular characteristics that do not make them representative of new or younger vegans.
"No method is perfect and I understand the criticisms of ours," says Anderson, "both the strict definition of diet on the one hand and including dietary instead of lifestyle vegans on the other. Both of those decisions would tend to produce recidivism rates on the high end.
"And studies like EPIC-Oxford that use self-identification will tend to go the other way," she continues. "It doesn't make one study more correct than the other unless you believe that one definition of vegan or vegetarian is absolutely correct and the other is absolutely false. It just provides a range of values, depending on how you think about those concepts.
"It's also worth noting of the EPIC-Oxford Study that they included only people over 35, thereby missing the largest target audience for veg*n advocacy."?
#4: Ask the wrong questions…?
So that might explain why the Faunalytics study produced such a startling result. But are there others?
The animal philosopher Vinciane Despret, author of What Would Animals Say if we Asked the Right Questions? could tell us: that the answers you get depend on the questions you ask. Looking at the Faunalytics survey might explain the figures their research provided.
The original Faunalytics research aimed to explore the reach of veganism and vegetarianism amongst the general U.S. population. When looked at from this angle, the research provided some real positives, including:
- Nearly 30 million people in the U.S. have tried vegetarianism or veganism (roughly one out of every eight people);
- vegetarianism has a positive reputation;
- people are more successful sticking to vegan diets than weight loss diets or attempts to quit smoking;
- More than three quarters of former vegetarians/vegans had no concerns about the impact that their veg*n diet was having on their health.
But examining the questions asked shows a potential bias toward recidivism and negative associations of veganism and vegetarianism.
Once the initial dietary question identifies and separates current omnivores from current vegetarians and vegans, the omnivores – currently eating meat products – are asked if they have ever eaten a vegan or vegetarian diet.
The survey then asks a range of questions about these dietary attempts. For example, after asking about the motivations for trying the diet the next question asks: "To what extent do you agree/disagree that the following statements applied to you when you ate a vegetarian diet?" Answering on a scale from strongly agree to strongly disagree, the options are:
- I found it complicated to prepare vegetarian meals
- I found it time consuming to prepare vegetarian meals
- I had trouble finding restaurants where I could eat/food I could grab on the go
- (R) I had easy access to a health food store or a grocery store with a health food section
- I had to prepare both vegetarian and non-vegetarian meals
- I had to rely on someone else to do the grocery shopping
- I found my vegetarian diet difficult during a time of transition (changing residences, traveling, starting a new job, etc.)
That R denotes a reverse of the scoring, from agree to disagree. But this illustrates how the question offers only one positive option to a neutral question; the rest imply negative responses to vegan/vegetarian diets, which frames the way the question is answered.
The question “To what extent do you agree/disagree that the following statements applied to you when you ate a vegetarian diet?” is limited to these significantly negative options:
- I craved/was tempted by beef or pork
- I craved/was tempted by poultry (chicken, turkey, etc.)
- I craved/was tempted by fish/seafood
- I was bored with my food options
- I was a picky eater
- (R) I found sufficient comfort/pleasure in vegetarian foods
- I felt deprived at restaurants or parties
So,this research overwhelmingly asked questions of current omnivores about their problems with a dietary choice that only a third of them trialled for less than three months, and another 53 percent for less than a year.
Compare this to the EPIC-Oxford Study, where self-identified vegetarians and vegans were asked if they were still following their chosen diets. After five years, 85 percent were. After 20 years, 73 percent were still either vegetarian or vegan.
But studies of self-defined diets are often also problematic. Juan et al. found that 27 percent of self-identified vegetarians had eaten red meat within the past 24 hours, and almost half (48 percent) had eaten meat if you include poultry and seafood.
"Those figures are astonishingly high," says Anderson, "so it's unsurprising that studies [like EPIC-Oxford] that use a self-definition of vegan or vegetarian find a much lower rate of lapsing than ours did: if eating a steak doesn't count as lapsing, it's pretty hard to lapse!"?
Was it the media?
Why we still remember, use, and mis-use, the 84 percent figure has a lot to do with how well and widely it has been communicated, and how easy it is to remember – and, to be fair, how it may chime with common sense understandings of how people change. And in its reporting, we know the media often picks up on the most controversial element of every story.
"It's a concern that applies to all research," agrees Anderson. "We try to guard against it by anticipating potential misinterpretations or over-interpretations and explicitly cautioning against them in our reports and by responding thoroughly to and encouraging questions from media and advocates."
But it wasn't just the media spinning the story this way. The language used by Faunalytics is explicit in counting those who've tried vegetarianism or veganism only once and even for less than three months as vegans or vegetarians. They even say quite plainly in the research findings: "84 percent of vegetarians/vegans abandon their diet."
But as some argue, trying a diet for less than three months does not ‘make’ you a vegetarian or vegan, and so to claim that would seem to be overstepping the mark.?
Answers to fit the agenda?
The research was conducted in 2014 by the then-Humane Research Council, which then rebranded as Faunalytics.
The key findings of the research suggest that concentrating on reduction rather than elimination, and adherence rather than adoption, might be more effective strategies for animal advocacy. This was in line with their long-held position on 'incrementalism'.
"Faunalytics has encouraged inclusiveness in the movement over the past five years, and we aren't alone in that," says Anderson. "With the rise of reducetarianism, Meatless Mondays, the call for reduced meat consumption to combat climate change, etcetera, the movement (including researchers) need to include and encourage people who are reducing."?
Shaped by Effective Altruism?
In their 2016 review of Faunalytics' work, Animal Charity Evaluators identified where this specific research on recidivism had shaped the strategies of a number of leading animal advocacy organisations.
For the past few years, the influence of Effective Altruism on the animal rights/advocacy movement, especially in the United States, has been huge. EA advocates using statistical analysis and measurable impacts to direct where individuals and organisations should place their philanthropy and funds.
Animal Charity Evaluators is an organisation based on EA principles. But the problem with this approach, as argued by many especially from an ecofeminist and care ethics perspective, is that a) not everything can be measured, and b) ACE are sometimes very wrong.
With organisations such as Faunalytics keen to achieve the ACE seal of approval (they've been a Standout Charity since 2015) and be worthy of philanthropic donations, research that fits the philosophical approach of Effective Altruism – against absolutism, with high-impact research statistics that go viral and are picked up widely by other organisations and in the media – are exactly what Faunalytics itself benefits from.
However, as the founder of Faunalytics, Che Green, has already pointed out, their position on incrementalism and on the critical importance of evaluative methodology for effective animal advocacy has been part of their ethos from 2000. This creation of a “culture of evaluation” therefore predates ACE by a long way.
Back in 2015, Green accepted the many criticisms of the report, and Faunalytics have introduced a number of improved procedures, including instituting policies of public pre-registration and open commentary for all major studies, such as sharing research plans for their major studies on the Open Science Framework.
A problem of representation
The use of phrases such as 'almost a third (30 percent) of former vegetarians/vegans had a lapse from vegetarianism/veganism prior to their present abandonment' could suggest a weak-willed population of wildly fluctuating do-gooders giving in to the sore temptations of animal products. Is this representation, or the research (or human nature!)?
"Methodological decisions impact results – there’s no way around that – so we consult with other researchers during the study design phase to ensure that we choose methods that are broadly supported and present them in a way that is as fair, unbiased, and clear as we can," says Anderson.
"And if a misinterpretation does unfortunately arise, we do what everyone else does and try to correct the record. This is an area in which Faunalytics has, in my opinion, improved over time and we continue to do so."
Others working in the field of vegan representation hope that Faunalytics’s new planned research will radically alter the language used around uptake and recidivism.
"The research problematically frames the issue of veg*nism as some kind of conversion on the model of (Christian) religious practice, by way of the terminology around 'lapsing'," says Dr. Bob McKay, a Senior Lecturer in English Literature at the University of Sheffield and a leading figure in the study of the representation of veganism and human-animal relationships.
"This is a symptom of a larger issue," continues McKay, "which is individualising, personalising, and so consumerising food ethics and dietary practice. This way of characterising the results obscures the difficulty of maintaining a healthy and ethically committed diet in the face of life pressures, systematic resistance, and the huge array of incentives, fiscal and otherwise (e.g. dairy subsidies, lobbying) waged against doing so."
"It's a wonder, surrounded as we are by animal products, that anyone manages to stay vegan at all," says long-time vegan activist and author of Growl, Kim Stallwood. (Kim celebrated 43 years as a vegan on January 1.)
If we were using the EPIC-Oxford Study figures of 27 percent, a quarter of people who had lived for a considerable time following vegan or vegetarian diets were, after twenty years, unable or unwilling to do so. Can we say why? As Anderson knows, more research is needed.
Yet that still approaches the numbers from the wrong end, argues McKay.
"A better tagline than 84 percent of veg*ns lapse," he continues, "just might be 'V are the champions: Against all odds, 16 percent of veg*ns win battle to maintain physical and moral health'."
And yet how much better does that sound if it’s not 16 percent but 85 percent or 73 percent of vegans and vegetarians who have, in carnist societies, maintained their ethical choices as measured over a period of 20 years?
I hope Piers Morgan is choking on his sausage roll at that one.