To determine the good done by convincing someone to change their diet to one that is better for animals, we need to know how long the changes will last. We investigate the general vegetarian recidivism rates in order to estimate the duration of the effects of any particular intervention.
- “Self-Reported Vegetarians” Really Means “Animal-Product-Limiters”
- Most Vegetarians Become Former Vegetarians
- Estimates Based on Former Vegetarians
- Estimates Based on Current Vegetarians
- Better Estimates
- Areas for Further Research
Many organizations focused on helping farmed animals do so at least in part through asking institutions or individuals to reduce consumption of animal products. In the case of institutional changes, usually relatively few institutions are involved and institutional policy shifts are easily documentable. This makes it reasonable to determine how long the change lasts through checking in with the institutions involved on an individual basis.
However, in the case of individual changes, determining the average length that a change lasts for participants in a particular program would require large scale surveys conducted over many years. (Or, because individuals’ reports of their own food consumption can be unreliable, an even more expensive program of objective tracking (using independently verifiable information like sales patterns at school cafeterias), again lasting for years.) While performing some follow-up studies is clearly beneficial for understanding the impact of a specific program, no program-specific studies now exist that track results over an extended period of time, and it is unlikely that conducting such studies for every program or type of program would be cost-effective. Instead, we use general research on vegetarianism to estimate the long-term effects of programs directed at convincing individuals to change their diet.
Several studies exist focused specifically on vegetarian recidivism, the phenomenon of people eating a vegetarian diet at one time and then returning to an omnivorous diet. Most of these studies do not explicitly address other animal-product-limiting diets such as veganism, pescetarianism, or meat reduction. However, many have tacitly addressed these issues by recruiting self-identified vegetarians and former vegetarians. Self-identified vegetarians often do not fit the strict definition of vegetarian but generally do limit their animal product consumption in some way. Thus they have addressed many dietary habits, but without distinguishing between them. Also, studies that address vegetarian recidivism typically have not addressed the original motivations for becoming vegetarian and so the effects they show are not distinguished by the original cause of a person’s vegetarianism1. Therefore, our analysis will be unavoidably simplified.
The best information we have about the actual dietary habits of self-reported vegetarians comes from an analysis of data from the Continuing Survey of Food Intake by Individuals (CSFII), conducted by the US Department of Agriculture. This survey asked a large representative sample of Americans whether they identified as vegetarians, and on separate occasions asked detailed questions about what they had eaten in the past 24 hours. Of those who identified as vegetarians, 64% had eaten what the study considered a non-negligible amount of meat in one or both 24 hour periods. However, even the self-reported vegetarians who had eaten meat had eaten significantly less of it on average than the non-vegetarians had (about 74% as much, by weight), and the meat they ate was significantly more likely to be fish, suggesting that at least some may have technically been pescetarians.
Self-reported former vegetarians may also be, to some extent, animal-product-limiters. One study2 found that they reported following a variety of eating patterns that reduced meat consumption without excluding it completely. Another study3, however, did not find former vegetarians’ diets differed significantly from non-vegetarians’ in terms of animal consumption. Based on existing data, it is not possible to conclude that former vegetarians’ diets differ meaningfully from those of people who have never been vegetarian.
Dietary studies that rely on self-reports can only give us a general idea of a person’s true diet, especially if the self-reports are claims of identity categories like “vegetarian” and “vegan” rather than reports of the actual foods eaten. Studies that ask for self-reported vegetarianism or former vegetarianism should be taken to refer broadly to animal-product-limiting diets, unless the study explicitly says it omitted any self-reported vegetarians that also eat meat, such as a recent study by Faunalytics. Faunalytics asked whether participants identified as vegetarian or vegan and also asked them to report what food they currently consume. This two-tiered approach allowed Faunalytics to omit any self-reported vegetarians or vegans that also ate meat. Readers of the Faunalytics study can be fairly certain that participants in the vegetarian and vegan categories do not consume meat, but this cannot be generalized to all studies reporting about vegetarians and vegans.
If most people who adopted a vegetarian or animal-product-limiting diet kept it up for the rest of their lives, the analysis of vegetarian recidivism might not be necessary. We could use well-established life expectancy figures to determine how many years of change to expect for each person who made a dietary change in the first place. However, this is not the case.
Studies suggest that the majority of vegetarians return to eating meat after a relatively short time. For instance, the Faunalytics study suggests that the number of former vegetarians4 in the US is five times higher than the number of current vegetarians. In their representative sample, 10% of respondents described themselves as former vegetarians, whereas only 2% were current vegetarians. In addition, CBS conducted a poll of a representative sample of the US population and found three times as many people described themselves as former vegetarians than as vegetarians. Although the margin of error of the poll allows for substantial variance in the ratio, it is likely that the number of former vegetarians in the United States is higher and closer to Faunalytics' estimates for two reasons. First, people who were vegetarian for a short time several years before the survey may not have recalled it at the time of the survey, whereas people who were vegetarian at the time of the survey almost certainly did recall this, even if they were only vegetarian very briefly. Second, some of the people who reported being vegetarian at the time of the survey would not remain vegetarian for their whole lives, and the survey did not have any means of detecting this.
Overall, the evidence suggests that the vast majority of people who adopt dietary changes like vegetarianism, veganism, and other forms of animal-product-limiting eventually return at least some distance towards their former dietary habits. An intervention that changes an individual’s diet usually has only a temporary effect. To know the size of the effect, we need to know how long it lasts.
The easiest way to find out how long vegetarians typically remain vegetarian is to ask former vegetarians how long they were vegetarian for before they went back to eating meat. Since former vegetarians know how long they were vegetarian, they can accurately report it (at least within the limitations of accuracy of any self-report). Several studies have done this, but since each of them (except the Faunalytics study) specifically recruited former vegetarians, it’s likely they gathered non-representative samples of the total population of former vegetarians. However, since they recruited through different methods, these errors may cancel out somewhat when the estimates are taken together. Two studies5 reported mean lengths of vegetarianism for their samples of former vegetarians, which were 9 and 3.3 years. A third study6 did not collect exact lengths of time of vegetarian diets, but reported a median time of 3-5 years. Our estimate for the average was 4.5 years. Faunalytics' study also did not collect exact lengths of time but reported a median time of 4-11 months (see Figure 1). Our estimate for an average time based on Faunalytics' data was 2.8 years (+/- 0.40 years).
These estimates may err in several ways. First, people who were vegetarian for only a short time, especially who were vegetarian longer ago, may have been particularly unlikely to respond to recruitment attempts from any study. Second, respondents may have had a systematic bias to report longer or shorter periods than they really were vegetarian for, or to pick round numbers (which, depending on the actual distribution, could have biased the group estimate up or down or not at all). Finally, these are not estimates that directly speak to the question of how long the average person remains vegetarian, because some people may remain vegetarian for all or nearly all of their lives, and these people are either impossible or extremely difficult to find in a sample of former vegetarians.
Three of the studies7 mentioned above also surveyed self-identified vegetarians on how long they’d been vegetarian. While current vegetarians don’t know how long they will remain vegetarian, they can at least say how long it’s been so far. This is a useful lower bound, especially given that in all cases the mean times produced were longer than the mean times former vegetarians had been vegetarian. Specifically, one found a mean time of 9.7 years (vs 3.3 for former vegetarians). A second had a median time interval of 6-10 years, and our estimate for the average was 6.3 years (vs 3-5 and 4.5 for former vegetarians). The third study, from Faunalytics, reported a median of over 10 years (see Figure 1) and our estimate for the average was 14.1 years (+/- 1.5 years, vs 4-11 months and 2.8 years (+/- 0.40 years) for former vegetarians).
Although these numbers are already above those produced for former vegetarians, we know they’re lower than the reality. It would be valuable to get estimates that were closer to the real average times these participants would remain vegetarian. One way to do so is to assume not that they’ve talked to the researchers on the last day they’ll be vegetarian, but on average about halfway through their period of vegetarianism. In a survey drawing randomly from the population, this would be a realistic assumption, though because these studies used nonrandom recruiting methods, it may not apply as well to them. In that case, the average length of vegetarianism would be about twice what the studies were able to observe -- 19.4, 12.6, and 28.2 years, respectively.
These estimates share all the problems of the estimates derived by talking to former vegetarians, except that instead of being unlikely to count people who are vegetarian for a long time, these estimates were especially unlikely to count people who are vegetarian for only a short time. Such people might be less likely to respond to recruitment efforts and also would only have a tiny window in which to do so as current vegetarians, so would be very unlikely to be involved in these studies.
Ideally, an estimate for the mean length of time vegetarians stay vegetarian would follow a group of new vegetarians until none were left, recording how long each had remained vegetarian. However, this would be expensive. Another possibility would be to randomly poll the general population, asking each person whether they were or had ever been vegetarian, and if so for how long. Luckily, Faunalytics did something similar to this by polling a representative U.S. sample and asking vegetarian respondents if they have been (or were) vegetarian for 0-3 months, 4-11 months, 1-2 years, 3-5 years, 6-10 years, or more than 10 years. By combining data from former and current vegetarians, we obtained an overall average length of vegetarianism of 7.03 years (+/-0.80). We believe that this is the strongest estimate we have since it is derived from data that is representative of the U.S. population and a process that strictly defined vegetarianism as zero meat consumption.
A similar estimate of 6.2 years can also be obtained by patching together estimates from other studies, but since these estimates have considerable uncertainty due to unrepresentative samples, smaller sample sizes, and unknown diet specifics, we would find any value between 3.6 and 13.3 years defensible from this data.
The existing data on vegetarian recidivism is incomplete. A long-term cohort study that follows vegetarians from the beginning to the end of their diet would be the best data for reliable population estimates. However, since this method would be expensive, a random sample from the population that asks for specific time lengths of vegetarianism would be a good alternative. Studies so far have also not reported on any differences in retention between self-reported vegetarians with different diets, and information on that topic would likely be immediately useful to groups working on creating change in individuals’ diets. Finally, our composite estimates for the mean length of vegetarianism, while useful, were ad hoc and lacking in rigorous statistical meaning; an expert statistician could revisit the data we have used and perform an analysis that would be more precise about the inferences that could be drawn.
Based on the available data, we believe that the average new vegetarian (or other animal-product-limiter, including vegans, pescetarians, and with less certainty adherents to meat reduction diets such as Meatless Mondays or Vegan Before 6) will retain their new dietary pattern for 7.03 years8 (+/- 0.80 years and a median of 1 year). It is possible but not substantiated that even after quitting their animal-product-limiting diet, former limiters will continue to eat somewhat fewer animal products than they had originally. Because the one study that included analysis based on participants’ motivations for vegetarianism did not find substantial differences in recidivism based on differences in motivation, it is likely that these findings are broadly applicable to any programs with similar target audiences, once differences in the rate at which people are convinced to go vegetarian are taken into account. Further research on a variety of points would be useful for reducing uncertainty and confirming that these results apply to the general population. However, we see limited value in multiple individual programs establishing the length of time that participants retain new dietary habits, because there is little evidence that the original cause of vegetarianism affects recidivism.
ACE. (2013). Vegetarian Recidivism statistics.
ACE. (2015). Faunalytics Data Analysis.
Alfano, S. (November 20, 2005). How and where America eats. CBS News.
Asher, K. et al. (2014). Faunalytics Study of Current and Former Vegetarians and Vegans.
Barr, S. I., & Chapman, G. E. (2002). Perceptions and practices of self-defined current vegetarian, former vegetarian, and nonvegetarian women. Journal of the American Dietetic Association, 102(3), 354-360.
Haddad, E. H., & Tanzman, J. S. (2003). What do vegetarians in the United States eat? The American Journal of Clinical Nutrition, 78(3), 626S-632S.
Haverstock, K., & Forgays, D. K. (2012). To eat or not to eat. A comparison of current and former animal product limiters. Appetite, 58(3), 1030-1036.
Herzog, H. (June 20, 2011). Why do most vegetarians go back to eating meat? [Blog post].
One non-representative retention study found minimal differences in vegetarian retention after three years between vegetarians who said that their primary reason for vegetarianism was health and vegetarians who said that their primary reason was ethics or animal rights. Other responses were available, but too few people selected them to draw any conclusions. Stahler, C. (2009). Do vegetarians and vegans stay vegetarian? The Vegetarian Resource Group. ↩
[T]he majority of former limiters were occasional meat eaters, followed by regular meat eaters, meat avoiders, and pescatarians." Haverstock, K., & Forgays, D. K. (2012). To eat or not to eat. A comparison of current and former animal product limiters. Appetite, 58(3), 1030-1036. ↩
Compared to nonvegetarians and former vegetarians, who were similar, fewer vegetarians consumed any flesh foods at least weekly." Barr, S. I., & Chapman, G. E. (2002). Perceptions and practices of self-defined current vegetarian, former vegetarian, and nonvegetarian women. Journal of the American Dietetic Association, 102(3), 354-360. ↩
Faunalytics reported findings from current and former vegans as well as larger numbers of current and former vegetarians. From this point on, we refer to the combined groups of vegans and vegetarians interviewed in that survey as current or former vegetarians for brevity. Faunalytics' initial findings report most statistics for these combined groups. ↩
Respectively, Herzog, H. (June 20, 2011). Why do most vegetarians go back to eating meat? [Blog post]. and Barr, S. I., & Chapman, G. E. (2002). Perceptions and practices of self-defined current vegetarian, former vegetarian, and nonvegetarian women. Journal of the American Dietetic Association, 102(3), 354-360. ↩
Haverstock and Forgays asked respondents to select the time range corresponding to the length of time they'd been vegetarian. Haverstock, K., & Forgays, D. K. (2012). To eat or not to eat. A comparison of current and former animal product limiters. Appetite, 58(3), 1030-1036. ↩
Barr, S. I., & Chapman, G. E. (2002). Perceptions and practices of self-defined current vegetarian, former vegetarian, and nonvegetarian women. Journal of the American Dietetic Association, 102(3), 354-360. and Haverstock, K., & Forgays, D. K. (2012). To eat or not to eat. A comparison of current and former animal product limiters. Appetite, 58(3), 1030-1036. ↩
Until March 2015, we used the estimate of 6.2 years (with error bounds of 3.6 and 13.3 years) obtained from the studies available prior to the release of Faunalytics' findings. Those figures are still reflected in calculations elsewhere on our site, though as we update other materials they will be replaced by the newer estimate. ↩