The sharing economy arrived with an appealing sustainability story. If idle cars, tools, and spare rooms could be accessed rather than owned, fewer products would need to exist. Utilization would rise, manufacturing would fall, and convenience and conservation would finally point in the same direction.

The logic holds on paper. In practice, many platforms have produced the opposite effect. Ride-hailing has added vehicle miles to congested cities. Dockless scooters have been discarded within months. Short-term rentals have made travel cheaper and more frequent. The problem is not that sharing is inherently wasteful. A system-level change was evaluated as if it were a product-level one.

From a life cycle perspective, sharing changes three variables at once: how much people consume, how intensively goods are used, and how quickly those goods wear out. Each can move in an unfavourable direction. Understanding these dynamics shows when sharing genuinely reduces resource use, and what designers, operators, and policymakers must get right for that to happen.

Economists have a name for what happens when efficiency makes something cheaper: the rebound effect. When the cost of a service falls, people use more of it. Sharing platforms reduce more than price. They also reduce friction, meaning the effort of planning, buying, storing, and maintaining. Something that once required a trip to a shop and a significant outlay now requires a tap. Lower friction is one of the strongest drivers of demand.

Ride-hailing shows the pattern clearly. Several studies of US cities have found that a substantial share of ride-hail trips replaced public transport, walking, or cycling, or would not have been made at all. Drivers also spend miles cruising between passengers, often called deadheading. Add those miles, and the net effect in many cities has been more vehicle travel, even as some households gave up their cars.

The same dynamic appears elsewhere. Short-term rental platforms lowered the cost and expanded the supply of accommodation, enabling additional trips and the flights that come with them. Fashion rental services can encourage more frequent wardrobe turnover, with each garment shipped, cleaned, and repackaged between wearers. In each case, access reshapes behaviour. Behaviour then determines environmental outcomes far more than the ownership model does.

The key analytical question is not whether a shared item replaces a purchased one. It is what the counterfactual looks like: what would people have done without the platform? When sharing substitutes for higher-impact activities, such as owning a second car, it delivers real gains. When it substitutes for lower-impact options or creates entirely new activity, total consumption rises even as each transaction looks efficient. Assessments that skip this step tend to overstate benefits considerably.

Takeaway

Efficiency per transaction says little about total impact. Before crediting a sharing model with savings, ask what activity it actually displaced and what new activity it created.

Sharing's central promise rests on underutilization. Private cars sit parked roughly 95 percent of the time. The household power drill, according to a statistic repeated in countless pitch decks, is used for only minutes across its entire lifetime. The implication is compelling. If one shared item could serve ten users, nine items would never need to be manufactured.

Real-world utilization rarely matches these projections. Demand clusters in time and space: everyone wants a car on Friday evening, a ladder on weekend mornings, a scooter downtown at rush hour. To stay reliable, operators must oversupply, which leaves much of the fleet idle off-peak. Peer-to-peer tool rental has often struggled with low transaction volumes, because borrowing still involves friction that owning does not. Marketing figures tend to reflect peak or theoretical use. Fleet-wide averages usually tell a more modest story.

Higher utilization also carries hidden costs. Shared assets require logistics that owned goods do not: rebalancing, collection, charging, cleaning, and inspection. Life cycle assessments of early dockless e-scooters found that manufacturing and the vans used to collect and charge them dominated their footprint. Early deployments also reported lifespans of only months. Under those conditions, many scooters produced more emissions per kilometre than the trips they replaced.

This does not invalidate the utilization argument. It refines it. The relevant metric is not how often an item is used, but impact per unit of service delivered across its full life cycle. A shared car that serves more trips is only better if those trips displace others, and if the overhead of sharing does not erase the manufacturing savings. Utilization is a means, not an end, and it can be optimized in ways that worsen environmental performance.

Takeaway

Utilization is only a proxy for sustainability. What matters is how much environmental cost each unit of delivered service carries once logistics, wear, and lifespan are counted.

Most products in sharing systems were originally designed for private ownership. A consumer scooter assumes a careful owner, occasional use, and sheltered storage. Placed on a public street and ridden by dozens of strangers a day, it breaks quickly. When products wear out faster than ownership would have required, the sustainability case for sharing collapses. Higher utilization simply compresses the same life cycle into less time, with added logistics on top.

Shared use demands a different design brief. Products must withstand intensive duty cycles, rough handling, and constant weather exposure. Later generations of shared scooters show the shift. Reinforced frames handled heavier use. Swappable batteries removed the need to collect whole vehicles for charging. Modular components could be replaced individually instead of scrapping the unit. Operators reported substantially longer vehicle lifespans as a result, which transformed the life cycle profile of the service.

This connects directly to cradle-to-cradle thinking. Michael Braungart argued that products should be designed as technical nutrients, cycling through use and recovery without losing value. Sharing platforms are well positioned to apply this principle, because the operator keeps ownership. A manufacturer selling to consumers can benefit from replacement sales. An operator profits from durability, repairability, and component recovery, so its incentives align with longevity.

Effective design for sharing typically includes robust materials sized for high-duty cycles, standardised and modular parts, diagnostics that enable predictive maintenance, and end-of-life pathways planned from the outset. Product-as-a-service models in commercial lighting and fleet tyres show how retained ownership encourages manufacturers to engineer for performance over time. Sharing becomes sustainable when the product is built for the system it serves, rather than adapted to that system after the fact.

Takeaway

A shared product is a different product. Sharing only reduces resource use when durability, modularity, and recovery are engineered in from the start, not expected from goods designed to be owned.

Sharing platforms are not inherently sustainable or unsustainable. They are system interventions, and their outcomes depend on how demand, utilization, and product design interact. Evaluating them requires full life cycle accounting and an honest counterfactual.

Three conditions tend to separate genuine gains from rebound. Sharing must displace higher-impact consumption rather than generate new activity. Operational overhead must stay small relative to manufacturing savings. And products must be designed for intensive shared use, with durability and recovery built in.

For practitioners, the guiding principle is straightforward: measure impact per unit of service, not per transaction or per user. A model that delivers more value from fewer materials, over longer lifespans, is sharing in the sense that actually matters.