Quantifying The Small Body Problem: A Meta-Analysis Of Animal Product Reduction Interventions

By JLRiedi @ 2025-09-03T22:18 (+18)

Do meat reduction campaigns risk harming more animals? A new meta-analysis from Faunalytics and Bryant Research finds no overall evidence that meat reduction campaigns push people toward eating more chickens or fishes, but results vary by intervention. The report highlights choice architecture as a promising approach and calls for more targeted research.

Background

The farmed animal protection movement deploys diverse strategies, including interventions designed to encourage individuals to reduce, or entirely eliminate, their consumption of animal products. As animal advocates and organizations launch and refine these campaigns, it becomes crucial to strategize around unintended consequences.

Substitution of one animal product for another can be an undesirable outcome for animal product reduction interventions, and poses a particular problem when a smaller-bodied animal product is substituted in place of a larger one. When people reduce their consumption of large-bodied animal products (e.g., pork, beef) but compensate by consuming more small-bodied animal products (e.g., chicken, fish) rather than plant-based substitutes, animal suffering is heavily compounded. How so? This substitution impacts many more animal lives, given the number of individual animals needed to fill the same demand. For example, nearly 200 chickens are required to produce the same amount of meat gained from slaughtering a single cow. This phenomenon is known in the farmed animal advocacy space as “the small body problem” and has even received attention in popular media.

Concerningly, two of the most prevalent appeals used to get people to reduce their animal product consumption — environmental appeals and health appeals — focus heavily on reducing red meat consumption and may explicitly favor the uptake of other animal products, such as fish or chicken. In light of the small body problem, this presents a major animal welfare concern. However, it remains unclear how prevalent the issue of substitution is and how concerned animal advocates ought to be about it when implementing reduction/elimination interventions. If meat reduction efforts risk unintentionally pushing people toward more small-bodied animal consumption, and thus costing more animal lives, animal advocates must be mindful of (1) the factors that lead to small-body substitution and (2) the best practices for guiding consumers toward plant proteins instead.

To address current ambiguity regarding the small body problem, Faunalytics and Bryant Research conducted a meta-analysis to evaluate evidence of a substitution effect in interventions designed to reduce or eliminate animal products from participant diets. That is, we investigated whether consumption of specific animal products ever increases in response to an intervention, even if overall animal product consumption decreases. Although we looked for evidence of increases across all edible animal products (EAPs), we were most interested in small-bodied animal products (e.g., fish and chicken). We also used the available evidence to analyze the intervention characteristics that were associated with substitution effects, allowing us to provide recommendations to advocates.

Key Findings

1. More research is vital. The meta-analyses undertaken in this report had statistical limitations, as there simply were not enough studies at our disposal to run the most robust and informative tests. Moreover, the studies currently at hand are highly variable, making it difficult to draw strong conclusions. We highly encourage advocates to conduct more studies on possible substitution effects that can be incorporated into meta-analyses such as the ones we report here.
2. Overall, we found no evidence of the small-body substitution problem in response to animal product reduction interventions. On the whole, people do not appear to increase their consumption of small-bodied animals (e.g., chickens, fishes) when exposed to animal product reduction interventions. However, this comes with the caveat that studies were highly variable, meaning that some interventions did actually result in greater small-bodied animal consumption, even when the overall result — with all studies considered — was null.
3. Overall, people did not decrease their consumption of animal products (neither large- nor small-bodied) when exposed to animal product reduction interventions. This may be alarming to advocates, as it implies that current advocacy strategies are ineffective. However, this is slightly reductive. Findings indicate that positive and negative results across the studies are canceling one another out, producing an overall null result. With this in mind, specific intervention types that do show success should be prioritized (see the fourth Key Finding). Moreover, new, innovative approaches ought to be considered.
4. Despite no overall effect, some individual interventions did result in reduced animal product consumption. Choice architecture in particular was associated with reduced consumption of small-bodied animals. We recommend advocates incorporate choice architecture into their efforts whenever possible to maximize impact.

Conclusions

We found no evidence of the small body replacement problem — people swapping out large animal products in their diet (e.g., beef, pork) for small animal products (e.g., fish, chicken) — in the present analyses. On the whole, when exposed to an animal product reduction intervention, people do not increase their consumption of small-bodied animals. Based on existing evidence, animal product reduction interventions do not appear to be backfiring. That is, they do not appear to be unintentionally driving people to consume more small-bodied animal products (nor large-bodied animal products, for that matter). However, we caution animal advocates to take these findings with a generous pinch of salt, given the tremendous variability between individual studies.

These results illustrate a pressing need for more replication studies — both those that are direct replications of past work and those that are simply conceptual replications. At present, we are comparing apples and oranges, as indicated by the extremely high heterogeneity between studies in each of the presented meta-analyses. Moreover, specific types of interventions were sparse (e.g., interventions delivered via a university class session, module, or course; interventions lasting longer than a single exposure; and many more). These circumstances leave us unable to draw strong conclusions about overall outcomes. However, if more studies adopt similar methods, we can more easily assess overall trends. To assist in this endeavor, we provide a research roadmap to help you design, run, and report your own interventions.

Granted, a range of studies do already exist that have investigated similar intervention mediums (e.g., choice architecture, documentaries, lectures), appeals (e.g., environmental, health, animal welfare), and so forth. However, many of these cannot contribute to our understanding of the small body problem, even when findings are significant, as they measure meat consumption in aggregate. That is to say, we only know how participants’ overall meat consumption changed, not how their specific consumption of chicken or beef changed, for example. If future studies take a more granular measurement approach, we can track people’s eating habits across specific species and better understand the state of the small body problem.

Regrettably, due to the statistical limitations of the analyses, we are not able to advise advocates on which characteristics to avoid when crafting their interventions. However, based on present — albeit limited — evidence, we recommend implementing choice architecture into intervention designs. For example, dining halls, restaurants, and cafés might institute a plant-based “dish of the day,” or retailers might display plant-based meat equivalents directly next to their animal-based counterparts. These choice architecture “nudges” facilitate higher plant-based uptake, at least temporarily, without impinging on consumers’ free choice and thus without stirring up controversy. It is perhaps unsurprising, therefore, that choice architecture was the soundest predictor of lower small-bodied animal consumption in our analyses. At the very least, choice architecture is very unlikely to trigger people to consume more small-bodied animal products, and some evidence indicates it triggers people to consume significantly less. We consider choice architecture a safe bet regarding the small body problem and encourage researchers to conduct more studies with this strategy; it is only with more research that the true effectiveness (or lack thereof) of this strategy will come to light.

Lastly, we foresee applications of the present data beyond the small body problem. We have collected a large swath of animal product reduction/elimination studies and reported granular details about them, from their means, standard deviations, and sample sizes to the specific animal products measured to the types of interventions undertaken and their timelines. Animal advocates can freely access these data for their own specific purposes, whether they aim to hone in on large versus small animals or not. We view this as a valuable resource for the animal advocacy movement and encourage advocates to use it freely and creatively. Similar data collections exist (see Rethink Priorities and LIME). In fact, LIME even provides an interactive meta-analysis feature, allowing users to run a meta-analysis within the platform itself. However, at the time of writing, LIME does not make a distinction between different meat types, making investigations regarding small-body substitution currently impossible (though this is subject to change). Our public data and code script uniquely address this.