Is plastic really better for the climate? It depends on the grid, and on what you count

A question we hear on occasion is some version of: “Isn't plastic actually better for the climate than glass or paper?” Usually the question shows up after a LinkedIn post, a trade-group infographic, or an op-ed arguing that cutting plastic use simply shifts emissions somewhere else.

The honest answer is more demanding than either side usually allows. Under many conventional life cycle assessments, plastic really does come out lower-carbon than heavier alternatives. But that finding is narrower than it sounds, incomplete as a guide to environmental impact, and likely to weaken as grids decarbonize and material systems change.

What the studies actually say

The headline result from Meng, Brandão, and Cullen, published in Environmental Science & Technology in 2024, is real and worth taking seriously: across 16 applications spanning packaging, construction, automotive, textiles, and consumer goods, plastic had the lowest greenhouse gas footprint in 15 of them ^[1]. That does not prove plastic is broadly environmentally superior. It does show that, on greenhouse gases alone, plastic often performs better than people expect.

A 2025 systematic review by Dolci and colleagues reached a similar high-level conclusion for packaging LCAs.^[2] Alternative materials can outperform plastic in specific systems, but the answer is highly conditional. Returnable glass can require many effective reuse cycles before it catches up to single-use PET on climate, and the break-even point depends heavily on bottle weight, transport distance, washing energy, and return rates.

These are not industry talking points. They are peer-reviewed results from independent academic groups using standard LCA methods. The mistake is not taking those results seriously. The mistake is asking them to answer more than they were built to answer.

Why plastic often wins in standard LCA

The first reason is simple: mass. Lightweight materials enjoy an advantage almost everywhere in a conventional packaging LCA. Less mass usually means less material production, less freight burden, and lower handling costs across the chain. That is one reason PET so often compares well against single-use glass ^[1-2].

The second reason is process energy, and this is where aluminum becomes especially revealing. Primary aluminum is highly electricity-intensive, while recycled aluminum is a completely different climate material. According to the International Aluminium Institute, recycled aluminum uses about 95.5% less energy than primary production ^[6]. Hydro reports its low-carbon primary aluminum at below 4 kg CO₂e per kg, compared with a world average around 14.8, while its recycled Hydro CIRCAL product is lower still at 1.9 ^[7]. That spread is the point: “aluminum” is not one emissions profile. Pathway matters. Grid matters. Recycled content matters.

Plastic, by contrast, often looks favorable today because it combines low mass with a production pathway that still fits efficiently inside a fossil-fueled industrial system. But that is not the same as saying plastic is inherently climate-benign. It means the present system rewards it on the metrics conventional LCAs are best at counting.

The overlooked caveat

This is the part that often gets skipped in pro-plastic readings of the literature: plastic’s climate advantage is partly a story about the present grid, not a permanent property of the material

As electricity gets cleaner, metals and glass can decarbonize sharply because a large share of their burden is tied to electricity and heat. Plastic also benefits from cleaner electricity, but not to the same extent, because a meaningful share of fossil-plastic emissions remains tied to feedstock carbon and thermally intensive upstream chemistry. That is why recent decarbonization work on plastics does not treat grid cleanup as sufficient on its own. A 2025 Nature Communications study concluded that getting the plastics system to net zero requires not only cleaner energy, but also high recycling rates and substantial shifts toward biogenic feedstocks ^[4]. Earlier work by Zheng and Suh likewise argued that efficiency alone is not enough to bend plastic’s emissions trajectory ^[5].

That does not mean plastic cannot decarbonize. It can. But fossil-feedstock plastic has a harder floor than people often acknowledge. You can clean up the electricity, electrify more heat, and improve recycling. But if the polymer itself still begins as fossil carbon, some part of the climate burden remains structurally harder to remove than it is for materials whose main burden sits more heavily in electricity-intensive processing rather than hydrocarbon feedstock.

Has anyone actually calculated the crossover?

Not cleanly, at least not in a form that has become standard in the published packaging literature.

Meng et al. include future grid sensitivities, and those scenarios show that some plastic advantages shrink as electricity decarbonizes ^[1]. Zheng and Suh model broader plastic decarbonization pathways ^[5]. The more recent Nature Communications work reinforces the same direction of travel: plastics do not reach deep decarbonization through electricity cleanup alone ^[4]. But none of those papers gives a simple, published crossover curve that says: at this grid carbon intensity, this specific PET container is overtaken by this specific aluminum or glass container.

So we built a simple illustrative sensitivity model for a 500 mL single-use primary container. It is not a peer-reviewed packaging LCA, and it should not be read as a definitive market-average comparison. It is a transparent way to ask the question the literature leaves partly open: how much of PET’s current climate advantage is really about today’s grid, and when do alternative materials begin to overtake it under cleaner electricity and different production pathways?

For each material, cradle-to-gate emissions are modeled as:

where e is process electricity intensity (kWh/kg material), I is grid carbon intensity (g CO₂e/kWh), and β is non-electric emissions, including process heat, direct process emissions, and feedstock-related carbon ^[1].

Per-functional-unit emissions are then:

where m is container mass in kilograms.^[1]

The table below lists the assumptions used in the model. Some rows are anchored more directly to published industry or sector data than others. In particular, the inert-anode aluminum row and the lightweighted glass row should be read as transparent scenario assumptions, not settled commercial averages ^[5-8].

The qualitative result is the point. In our model, recycled aluminum looks very different from primary aluminum. Low-carbon primary aluminum looks very different from conventional carbon-anode aluminum. Single-use glass remains hard to justify on climate unless reuse is doing real work or bottle mass falls sharply. And PET’s advantage narrows as the grid cleans, but does not disappear simply because electricity gets cleaner.

That model also has clear boundaries. It isolates upstream material production rather than trying to solve the entire packaging system at once. Transport, end-of-life, and forming or conversion are excluded from the baseline. That keeps the comparison legible, but it also means the chart should be read as a transparent sensitivity analysis, not a final verdict on packaging policy.

The deeper problem with the question itself

Even then, climate is not the whole argument.

Standard LCA is built to count greenhouse gases, energy use, acidification, eutrophication, and related categories. It is not yet well built to count many of the harms that make plastic uniquely concerning: long environmental persistence, diffuse fragmentation into microplastics and nanoplastics, biological uptake, and downstream ecological consequences of that loading. Reviews of packaging LCAs note that plastic leakage and littering are still not routinely represented in standard life cycle impact assessment ^[3].

That omission matters more now because the microplastics literature is beginning to touch climate-relevant systems directly. A 2025 PLOS One paper showed that microplastic contamination can distort measurements used to understand the ocean’s carbon cycle, including enough in one modeled case to make a sample appear about 4,000 years older than it is ^[8]. A separate 2025 PNAS paper estimated potentially large photosynthesis losses under microplastic pollution across terrestrial and aquatic systems ^[9]. That paper is important, but it should still be handled carefully: outside experts have urged caution in treating those global estimates as settled ^[10]. Those are exactly the kinds of effects packaging LCAs were not designed to see.

So when someone says, “plastic is better for the climate than glass or paper,” the most honest response is: under many standard LCAs, often yes ^[1-2]. But that is not the same thing as saying plastic is the better environmental material overall, because the framework producing that answer excludes some of the very harms that now define the material scientifically and politically ^[3,8-10].

Where we stand

We accept the narrow LCA result. Under today’s still fossil-heavy grids, many plastic packaging formats do rank lower on greenhouse gases than heavier alternatives ^[1-2]. We do not think that finding supports the broader conclusion it is often used to support.

First, much of plastic’s current climate advantage is contingent, not permanent. As grids decarbonize, electricity-heavy materials can get dramatically cleaner. Fossil-feedstock plastic can improve too, but it does not decarbonize along the same curve without deeper feedstock change ^[4-5].

Second, the LCAs most often cited in plastic’s defense systematically omit emerging impact pathways that matter, including plastic leakage, microplastic persistence, and downstream effects on systems tied to carbon cycling and productivity ^[3,8-10].

Third, climate is one dimension of material choice, not the whole of it. A material can score well on greenhouse gases inside a standard industrial model and still perform badly on persistence, diffuse contamination, and internal human exposure. Those are not side issues with plastic. They are central issues.

Plastic’s climate advantage is real in many present-day comparisons. It is also partial, conditional, and measured with a framework that still misses some of plastic’s most distinctive harms. If we are going to talk honestly about materials, all of those facts need to sit on the table at the same time.