Carbon food labelling does shift purchasing decisions — but the effect is modest, context-dependent, and easily undermined by competing signals like price, taste, and convenience. Randomised controlled trials show reductions in the emissions intensity of food purchases of roughly 5–10% when carbon labels are present. The effect is stronger when labels use traffic-light formats and when high-emission items are clearly distinguished from lower-emission alternatives.
Carbon labelling on food is expanding. Oatly, Quorn, and a growing number of food manufacturers now display lifecycle emissions on packaging. Several UK and European retailers have piloted shelf-edge carbon labels. In the US, legislative proposals for mandatory food carbon disclosure have been introduced at the federal level, though none have passed as of early 2026.
The evidence base on whether these labels actually change behaviour has grown substantially over the past five years. This post reviews what the research shows — including where labelling works, where it doesn’t, and what it implies for individual decision-making.
What the trials show
The strongest evidence comes from randomised controlled trials conducted in cafeteria and simulated shopping environments. A 2021 study by Camilleri et al. in Nature Climate Change — the most cited recent trial — found that displaying carbon footprint labels alongside food options reduced the emissions intensity of choices by approximately 9.3% compared to a control group with no labels.
A 2022 meta-analysis by Goossens et al. covering 17 studies found a weighted average effect of approximately 6.5% reduction in purchase emissions intensity. The range was wide — from near-zero effects in some real-world supermarket pilots to effects above 15% in controlled laboratory settings.
Field studies in actual retail environments tend to show smaller effects than laboratory studies — a common pattern in behavioural research where real-world distractions, habitual purchasing, and time pressure attenuate the influence of information interventions.
What makes carbon labels more effective
Label format matters significantly. Research consistently identifies four design factors that increase effectiveness:
A label showing “High / Medium / Low” emissions in red, amber, and green produces larger behavioural effects than a label showing “3.2 kg CO₂e.” Numerical labels require the consumer to perform a comparative calculation that most people don’t do at point of purchase. Colour coding performs the comparison for them.
Labels that show a product’s emissions relative to a category average (“30% above average for this food type”) produce stronger effects than standalone absolute figures. People anchor on relative comparisons more readily than absolute ones. This is consistent with how nutritional traffic lights work — the reference point is the food category, not an abstract ideal.
Labels on shelf edges at point of selection are more effective than labels on packaging that is only visible after the item is in the basket. The timing of information presentation relative to the decision moment matters — post-selection information produces awareness but not behaviour change.
Carbon labels have no effect if consumers cannot act on the information — if all options in a category have similar emissions, or if lower-emission options are unavailable, the label produces awareness without purchase change. The effect is largest in categories with high emissions variance, like meat versus plant protein.
Where carbon labelling fails
The literature also documents consistent failure modes:
- Price dominance: When a higher-emission option is significantly cheaper, carbon labels rarely overcome the price signal. Effect sizes drop sharply when price differentials exceed 10–15%.
- Greenwashing dilution: When low-credibility carbon claims proliferate (vague “eco” branding without verified data), consumer trust in all carbon labels erodes — including credible ones. This is documented in UK retail surveys post-2022.
- Habitual purchasing: For staple items bought on autopilot — weekly grocery items, regular lunch choices — label effects are minimal because purchasing decisions are not being actively evaluated at point of selection.
- Voluntary adoption bias: Companies that voluntarily adopt carbon labelling tend to be those with lower-emission products. This creates selection bias in observational studies and overstates the likely effect of mandatory labelling across the full product range.
The methodology problem: whose numbers?
Carbon labels are only as credible as the methodology behind them. There is currently no mandatory standard for food carbon labelling in any major market, which means the figures on products can reflect very different scope and boundary assumptions.
Common inconsistencies include whether land use change emissions are included (they are large for beef and soy; many manufacturer-funded studies exclude or minimise them), whether packaging and distribution are in scope, and which GWP values are used for methane and nitrous oxide.
A credible carbon label should state: the system boundary (farm gate only, or cradle to grave?), whether land use change is included, the GWP values used, and the third-party verifier. Labels that don’t disclose these assumptions cannot be compared against each other or against independent datasets like Poore & Nemecek (2018). See how Decarb approaches emission factor transparency.
What carbon labelling means for your footprint
For someone trying to reduce their food emissions, carbon labels are useful — but not sufficient as a sole information source. Three practical points follow from the evidence:
- Use labels for within-category comparison, not cross-category. A carbon label is most useful for comparing two similar products — two types of milk, two ready meals. Cross-category comparisons (beef vs tofu) are better made using independent lifecycle data than manufacturer-supplied labels, which may not use consistent methodologies.
- Prioritise category-level decisions over brand-level ones. The emissions gap between beef and chicken is larger than the gap between any two beef brands. Optimising within beef using carbon labels produces smaller gains than simply reducing beef frequency. The research on high-impact actions consistently points to category switching, not brand switching, as the higher-leverage choice.
- A personal footprint estimate is a more reliable guide than point-of-purchase labels. A structured annual footprint calculation — covering all major emission sources — gives you a ranked picture of where diet fits relative to transport, energy, and other categories. That ranking is more actionable than individual product labels, which only address one food item at a time.
See how food fits into your full carbon picture
A full footprint estimate ranks diet against transport, energy, and goods — so you know where to focus first.
Frequently asked questions
Does carbon food labelling actually work?
Yes, but modestly. Randomised controlled trials show reductions in purchase emissions intensity of approximately 5–10% when carbon labels are present. Real-world field studies tend to show smaller effects. The impact depends heavily on label format, product category, and whether lower-emission alternatives are available.
Which carbon label format is most effective?
Traffic-light colour-coded formats (red/amber/green) outperform numerical labels in most studies. Labels that include comparative context — showing how a product compares to the category average — are more effective than standalone absolute figures. Placement at the point of decision (shelf edge rather than packaging) also increases impact.
Are carbon labels on food reliable?
Reliability varies significantly. There is no mandatory standard for food carbon labelling in any major market, so figures can reflect different scope assumptions — particularly whether land use change is included and which GWP values are used for methane. A credible label discloses its system boundary, GWP values, and third-party verifier.
Is carbon labelling mandatory anywhere?
As of early 2026, no major market has mandatory carbon labelling for food products. The EU Farm to Fork Strategy references carbon labelling as a policy goal, and several national-level proposals have been introduced, but none have been legislated. Current adoption is voluntary and concentrated among manufacturers with lower-emission products.
Should I use carbon labels to guide my food choices?
They are useful for within-category comparisons but less reliable for cross-category decisions due to inconsistent methodologies. For the highest-leverage food decisions — such as reducing beef frequency or shifting to plant proteins — independent lifecycle data from sources like Poore & Nemecek (2018) is more reliable than manufacturer-supplied labels. A full personal footprint estimate provides the most actionable picture.
Sources
- Camilleri, A.R. et al., “Consumers underestimate the emissions associated with food but are aided by labels,” Nature Climate Change, 9, 2019.
- Goossens, Y. et al., “The effect of eco-labels on consumer purchasing behaviour: a systematic review,” Journal of Cleaner Production, 2022.
- Thøgersen, J. et al., “Consumer responses to ecolabels,” European Journal of Marketing, 2010. Label format and trust.
- Poore, J. & Nemecek, T., “Reducing food’s environmental impacts through producers and consumers,” Science, 360(6392), 2018.
- European Commission, Farm to Fork Strategy, 2020. Carbon labelling policy context.
- Carbon Trust, Carbon Footprinting and Labelling: Lessons from the UK, 2023.
