The Functional Unit Decides the Answer: What a Grocery Bag Study Reveals About Comparing Anything

In short: Ask whether a reusable bag beats a disposable one and you'll get a confident answer from almost anyone. A Clemson University life cycle assessment shows the honest answer is a question: used how many times? — and that the way you frame that question determines the result before a single impact is calculated.

A study that measured behaviour instead of assuming it

Most bag comparisons start with an assumption: one reusable bag replaces some number of disposable ones. The number is usually asserted rather than measured.

Robert Kimmel and colleagues at Clemson University took a different approach in their 2014 Life Cycle Assessment of Grocery Bags in Common Use in the United States (Clemson University Digital Press). Before modelling any environmental impact, they ran a controlled experiment to establish how many bags people actually use.

The setup: 60 participants, four bagging lines, four bag types, and four identical sets of 52 grocery items representing one shopping trip for a U.S. family of four. Participants received general instructions from people with grocery bagging experience and could leave items unbagged if they judged a bag unnecessary. Bag order was randomised per participant.

The 52-item list was built from U.S. Census Bureau data: a family of four spending roughly $221 per week on food under a moderate cost plan, plus an estimated $80 on household and personal care items. Food Marketing Institute data indicated consumers made 2.2 grocery trips per week in 2012, of which about 1.85 involved bags.

What the measurement found

The results, per single 52-item trip:

| Bag type | Bags used | Total bag weight | |---|---|---| | Plastic retail bags (PRB) | 9.8 | 61.0 g | | Paper | 8.4 | 457.2 g | | Reusable LDPE | 8.3 | 295.6 g | | Reusable NWPP | 6.7 | 621.9 g |

The NWPP bag held the most items per bag; the thin plastic bag held the fewest. About four items per trip went unbagged regardless of bag type — milk jugs, pet food, paper towel packs.

The differences were tested by ANOVA and hold up: an F value of 24.21 against an F-critical of 2.64 at 95% confidence, with a very low p value. Where the bag types differ in average number used, those differences are real rather than noise.

One incidental finding worth noting for anyone designing packaging: the filled bags weighed roughly 3.2 to 4.9 kg, while the U.K. Environment Agency reported carrier bag weight capacity at 18 to 19 kg. Bag capacity is limited by volume, not by strength. Making a bag stronger doesn't reduce how many are needed. Making it hold more volume does.

The choice that determines everything

Here is where the study becomes instructive far beyond grocery bags.

Because the bags differ in weight, capacity, and durability, the authors defined four separate functional units rather than one:

Bags for 1 trip (52 items)
Bags for 3.1 trips (161.2 items)
Bags for 14.6 trips (759.2 items)
Bags for 44 trips (2,288 items)

Now watch what happens to the material required as the functional unit stretches:

| Bag type | 1 trip | 3.1 trips | 14.6 trips | 44 trips | |---|---|---|---|---| | PRB | 61 g | 189 g | 891 g | 2,684 g | | Paper | 457 g | 1,417 g | 6,675 g | 20,116 g | | LDPE reusable | 296 g | 296 g | 296 g | 296 g | | NWPP reusable | 622 g | 622 g | 622 g | 622 g |

At one trip, the disposable plastic bag uses the least material by a wide margin — 61 g against 622 g for the NWPP reusable. At 44 trips, the same comparison inverts entirely: 2,684 g against 622 g, and paper reaches 20,116 g.

The reusable bags' totals never change, because the same bags are reused. The disposables scale linearly.

This is the entire debate in one table. Anyone can produce a defensible-looking answer favouring either side simply by choosing where to stop counting. And that choice — the functional unit — is made before any environmental data enters the model.

Why this generalises beyond bags

The functional unit is the quantified performance of a product system used as a reference. It sounds like a technical formality. It is in fact the most consequential judgment in any comparative assessment, because it defines what "equivalent" means.

The pattern recurs everywhere durability and disposability compete:

A durable appliance versus a cheap one — over what service life?
Refillable versus single-use packaging — at what return rate?
On-premise infrastructure versus cloud services — over what refresh cycle?
Remanufactured versus new components — across how many cycles?

In every case, the honest answer depends on a usage assumption that sits outside the model. When a supplier presents a comparison showing their product wins, the first question isn't whether the calculation is right. It's what functional unit was chosen, and whether it matches how the product will actually be used in your operation.

The transparency that makes a study usable

Several methodological disclosures in this study are worth noting as a benchmark for what a defensible comparison looks like.

Recycled content was modelled explicitly, with both credit and burden. Following ISO 14044, recycling used the cut-off method: a credit where recycled material replaced virgin material, and a burden for reprocessing — collection, transport, cleaning, sorting, pelletising. Recycled content is not treated as free.

Assumptions were stated rather than assumed favourable. NWPP bags were modelled at 0% recycled content and 0% recycling at end of life, because non-woven fabrics foul sorting machinery, handles are difficult to process, and intense colours limit recycling options. The authors cross-checked this against a parallel study by Muthu and Li covering China, Hong Kong and Thailand, whose authors had local knowledge and made the same assumption.

Geography was modelled to real supply chains. PRB production was located in Indiana, LDPE in California, paper in Ohio, polymer feedstock near the Texas Gulf Coast, NWPP manufacturing in China with Chinese grid electricity and transoceanic freight. Distance from bag manufacturer to supermarket was set at 450 miles across all types.

Data age was disclosed as a limitation. Polymer data from 2008, paper and several inputs from 2003, extrusion from 2005, recycling updated to 2012. The authors state plainly that processes may have improved since, which would lower actual impacts.

Secondary use was separated from the base case. PRBs are frequently reused as bin liners; paper bags get reused too. Rather than quietly folding this benefit into the main result, the authors excluded it from the base case and treated it as an alternative scenario using system expansion, crediting the avoided manufacture of trash-can liners.

A critical review panel was convened, as ISO 14040 requires for comparative assertions intended for publication: Vee Subramanian (chair, PRé North America), Katherine O'Dea (GreenBlue), and Dr. Susan E. M. Selke (Michigan State University).

Comparison across studies — and why rankings drift

The authors compared their measured reference flows against four other studies. The ratio of plastic bags to paper bags used per trip:

Franklin Associates: 1.5:1 to 2:1
Boustead: 1:1 to 1.5:1
U.K. Environment Agency: 1.26:1
Chico: 1.5:1
Clemson (measured): 1.17:1

Same comparison, five different ratios. Bag sizes almost certainly differed across studies, and the authors say so. But the spread illustrates why comparing conclusions across LCA studies is hazardous: the inputs that drive the ranking are often established before the environmental modelling begins.

For scale on what's at stake: total U.S. consumption of new plastic retail bags runs to roughly 100 billion bags per year — about 1.4 billion pounds of material annually at 6.2 g per bag.

The practical takeaway

Two questions turn this into something usable.

First: what usage assumption is this comparison built on, and does it match reality? A reusable bag used twice and discarded is worse than the disposables it replaced. Used forty times, it isn't close. Neither number is wrong — they answer different questions. The same logic applies to any durable-versus-disposable decision in an operation.

Second: is the winning option winning on the metric that matters, or on the framing? When a comparison's result flips depending on where the counting stops, the result isn't a finding. It's a parameter.

For organisations building circular business models — refillable packaging, product-as-a-service, remanufacturing — the functional unit question isn't academic. It determines whether the model actually delivers the environmental benefit it promises, and at what usage rate the promise starts being true.

Frequently asked questions

What is a functional unit in life cycle assessment? The quantified performance of a product system used as a reference for comparison. It defines what "equivalent" means — and it is the most consequential judgment in any comparative assessment, because it is chosen before any environmental data enters the model.

Is a reusable bag better than a disposable one? It depends entirely on how many times it is used. In the Clemson study, one shopping trip required 61 g of plastic retail bags versus 622 g of reusable NWPP. Across 44 trips, the disposables reach 2,684 g while the reusable bags remain at 622 g, because the same bags are reused.

How many bags does a typical grocery trip actually require? Measured with 60 participants bagging 52 items: 9.8 plastic retail bags, 8.4 paper bags, 8.3 reusable LDPE bags, or 6.7 reusable NWPP bags. The differences were confirmed by ANOVA with an F value of 24.21 against an F-critical of 2.64.

Does making a bag stronger reduce how many are needed? No. Filled bags in the study weighed roughly 3.2 to 4.9 kg while carrier bag weight capacity is 18 to 19 kg. Bag capacity is limited by volume, not strength — so increasing volume reduces bag count, increasing strength does not.

Why do different studies reach different conclusions about bags? Because reference flows differ. The ratio of plastic to paper bags per trip was 1.17:1 in the Clemson measurement, 1.26:1 in the UK Environment Agency study, 1.5:1 in another, and up to 2:1 in a fourth. The inputs driving the ranking are established before environmental modelling begins.

How does this apply outside grocery bags? The same logic governs any durable-versus-disposable decision: refillable versus single-use packaging at a given return rate, remanufactured versus new components across a number of cycles, or durable versus cheap equipment over a defined service life.

How Sustek.co makes circular claims defensible

A circular business model that only works above a usage threshold nobody has calculated is a marketing risk, not a strategy. Sustek.co establishes the threshold — and the evidence behind it:

Sustainability Pulse — Maps where material impacts actually concentrate in your operation. The Sustrategize™ baseline diagnostic delivers ESG maturity assessment, data gap analysis, and circular economy potential by product line.

Sustainability Navigator — Builds the analysis into Circular Business Model Redesign: establishing the usage thresholds at which a circular model outperforms the linear one it replaces, documenting the assumptions so the claim survives scrutiny, and quantifying circular value-at-stake in figures the CFO recognises. Delivered as a board-ready Transformation Blueprint aligned to CSRD, GRI and TCFD.

Sustainability Command — Tracks actual performance against those thresholds through Quarterly Executive Impact Dashboards, with continuous social value measurement via ValueFlow™, powered by the MeasureUp proxy bank.

When a redesign needs expertise you don't hold internally — reusable packaging systems, return-rate modelling, extended producer responsibility in a specific jurisdiction — Iconet® activates a network of 93+ vetted sustainability consultants on demand, with no fixed structural cost. The Iconet® x AWS reStart program carries a verified SROI of 6.41:1.

Sustek.co is an S.A.S. BIC certified under Sistema B Colombia, and part of Google for Startups and AWS Activate.

Do you know the usage threshold at which your circular initiative actually beats the alternative? If the answer is an estimate rather than a calculation, start here: sustek.co/services

Source: Kimmel, R., Sc.D. (2014). "Life Cycle Assessment of Grocery Bags in Common Use in the United States," Chapter 3: Scope, pp. 27–32. Clemson University Digital Press. Reference flow study conducted at Clemson University with 60 participants. Modelled in SimaPro using the US-EI 2.2 database; impacts calculated via World ReCiPe Midpoint H/A V1.07, IPCC 2007 100-year V1.02, and Cumulative Energy Demand V1.08. Critical review panel per ISO 14040. Note: this article draws on the study's scope and reference flow chapter; impact results are reported in later chapters of the full publication.