Japan’s Used Car Exports Are Not Automatically Green, Landmark Study Finds

Every year, more than a million used cars leave Japanese ports bound for roads in Nairobi, Karachi, Ulaanbaatar and beyond. As wealthy economies race to electrify their fleets, a growing share of these exports are hybrids and electric vehicles, and the intuitive assumption has been that giving an old car a second life somewhere else must be good for the climate. A new study published in the Journal of Industrial Ecology by researchers at the University of Tokyo and Waseda University dismantles that assumption with unusual precision. By decomposing the entire post-deregistration life of an exported vehicle into separable modules, the team shows that the climate benefit of the used-vehicle trade is contingent rather than inherent: it depends overwhelmingly on whether the destination country can actually recover materials at the vehicle’s end of life, and only secondarily on how clean its electricity grid is.

The scale of the trade is enormous. In 2024 alone, Japan exported roughly 1.5 million used vehicles to markets ranging from Southeast Asia to Africa, and electrified powertrains rose from just 6 percent of that flow in 2015 to 21 percent in 2024. The United Nations Environment Programme estimates that at least 23 million light-duty vehicles were exported to the Global South between 2015 and 2022, yet 75 importing countries maintain weak or very weak regulatory frameworks. Earlier research has already documented how this trade can offshore emissions: a national-accounting study found that Japan’s domestic vehicle electrification reduced emissions at home while increasing them abroad through global supply chains, and work on the US-Mexico corridor identified a tension between a technique effect, where cleaner imports displace older polluters, and a life extension effect, where cheap imports simply prolong the total driving stock.

What has been missing, the authors argue, is a systematic comparison across multiple powertrains and multiple destinations that isolates which system-level drivers actually determine whether an export is a net climate benefit or burden. To fill that gap, they built a modular attributional life cycle assessment with avoided-burden modeling, implemented in openLCA with the ecoinvent v3.11 database and ReCiPe 2016 impact assessment. The functional unit was the post-deregistration life cycle of a single passenger vehicle, and the system boundary was gate-to-grave, starting at the moment the car was deregistered in Japan. Manufacturing itself sits outside the boundary and enters only as a life-extension credit: keeping a used car on the road defers the production of a replacement, and that deferral is counted as avoided manufacturing emissions.

The model split the outcome into four modules: international ocean transport, manufacturing life extension, second-life operation over an assumed 35,000 kilometers across five years, and end-of-life treatment, which was further resolved into a vehicle-body subsystem covering steel, aluminum, glass and plastics, and a traction-battery subsystem for plug-in hybrids and electric vehicles based on hydrometallurgical recycling of NMC111 chemistry. Three reference cases bounded the analysis: domestic recycling in Japan, export with second-life use and landfill-only disposal at the destination, and export with second-life use plus destination recycling with recovery credits. Ten major recipients of Japanese used vehicles were examined, including the United Arab Emirates, Russia, New Zealand, Mongolia, Tanzania, Chile, Kenya, South Africa, Malaysia and Pakistan, spanning grid carbon intensities from about 0.16 to 1.17 kilograms of CO2-equivalent per kilowatt-hour.

The headline result is stark. Without destination material recovery, exports were net climate burdens for every powertrain except electric vehicles on very low-carbon grids, and the absence of recovery raised mean net global warming potential by approximately 8,400 kilograms of CO2-equivalent per vehicle relative to domestic recycling. Under domestic recycling, every powertrain delivered a net benefit of between 6,700 and 5,100 kilograms of CO2-equivalent per vehicle, driven entirely by the avoided-burden credit for recovered materials. When that credit vanished at the destination, the balance flipped. The availability of end-of-life material recovery, rather than the export flow itself, therefore governed whether the system was climate-beneficial, and it ranked first among the system-level drivers, ahead of grid carbon intensity in second place and powertrain type in third. Transport distance, perhaps surprisingly, did not alter the ranking at all.

When destination recovery was included, hybrids, plug-in hybrids and electric vehicles were net-beneficial across all ten destinations, with net benefits ranging from 8,500 down to 140 kilograms of CO2-equivalent per vehicle. Electric vehicles delivered the largest benefit, from minus 8,500 kilograms per vehicle in New Zealand to minus 2,500 in South Africa, while conventional cars were destination-dependent, ranging from a 2,000-kilogram benefit in Tanzania to a 1,200-kilogram burden in Kenya. The recovery credit is concentrated in aluminum: for a hybrid, recovered aluminum contributed 3,733 kilograms of the total 5,665-kilogram recovery credit, because recovered aluminum was credited against primary ingot at 22.0 kilograms of CO2-equivalent per kilogram. On the NMC111 basis, the total recovery credit corresponded to 48 percent of cradle-to-gate manufacturing emissions for electric vehicles, 92 percent for plug-in hybrids, 97 percent for conventional cars and a striking 111 percent for hybrids, meaning the credited recovery exceeded the entire manufacturing burden of the vehicle.

The use phase revealed a second, equally consequential dynamic. For conventional cars and hybrids, whose fuel supply was modeled on a standardized global petrol chain, net use-stage burdens were destination-invariant at 4,142 and 3,021 kilograms of CO2-equivalent per vehicle respectively. For electric vehicles, the outcome hinged on the destination grid. In Kenya and New Zealand, where grid intensity is at or below 0.17 kilograms of CO2-equivalent per kilowatt-hour, the manufacturing life-extension credit exceeded cumulative operating emissions, yielding net use-stage values of minus 2,183 and minus 2,144 kilograms per vehicle, so the imported electric car delivered a climate benefit through life extension alone. The breakeven grid intensity was approximately 0.58 kilograms of CO2-equivalent per kilowatt-hour; above that threshold electric vehicles became net positive in the use stage, though they still outperformed conventional cars at every destination. Extending the second-use period from five to seven years amplified these effects in both directions, deepening electric vehicle benefits on clean grids to around minus 4,000 kilograms per vehicle while worsening outcomes for petrol powertrains, since two extra years of combustion more than offset the larger credit.

The sensitivity analysis exposed how fragile some of these conclusions are to methodological choices. Only the electric vehicle benefit was robust to the basis used to credit recovered aluminum. Hybrid exports, beneficial under region-specific primary and market-wrought aluminum credits, reversed to net burdens in all ten destinations when recovered aluminum was credited against secondary cast alloy. Under a cut-off allocation that withheld recovery credits entirely, only electric vehicle exports to Kenya and New Zealand remained net-beneficial. Battery chemistry, whether NMC111, NMC622 or NMC811, did not change the sign of the electric vehicle balance, and a screening test applying a 1.5 multiplier to use-phase electricity intensity turned electric vehicle exports into a net burden only in South Africa, the destination with the most carbon-intensive grid.

To translate these findings into governance terms, the authors classified destinations by readiness archetypes. New Zealand, combining a clean grid of 0.165 kilograms of CO2-equivalent per kilowatt-hour with established end-of-life governance, produced the largest electric vehicle benefit of all ten destinations. Kenya pairs a favorable grid with nascent recycling infrastructure, so electric imports are already climate-beneficial, but reaching parity with domestic Japanese recycling required recovery rates 1.6 to 1.7 times the modeled baseline because Kenya displaces relatively low-carbon primary aluminum. Mongolia, with a grid intensity of 1.173 kilograms of CO2-equivalent per kilowatt-hour and no formal end-of-life recycling framework, showed that even the largest modeled benefits materialize only when destination recovery is actually realized. The authors also flag transit and re-export hubs such as the United Arab Emirates, Chile and South Africa, where the nominal importer differs from the final country of use, arguing that effective governance must trace conditions at the point of actual operation rather than the port of first entry.

The study is careful about its limits. The life-extension credit assumes the displaced new vehicle shares the powertrain of the exported one, and the model does not capture rebound effects or induced demand. Recovered aluminum is credited against region-specific primary production while vehicles are built from market-mix alloys, an asymmetry the authors acknowledge inflates the aluminum credit. LFP battery chemistries were excluded, in-service battery degradation was not modeled, and the analysis focused on global warming potential, though a screening across all 18 ReCiPe midpoint categories confirmed the driver ordering was directionally robust while identifying mineral resource scarcity and toxicity as the most end-of-life-relevant indicators. Within those bounds, the message is unambiguous: the climate case for used-vehicle exports is not automatic. It requires destination material recovery for conventional cars, hybrids and plug-in hybrids, and for electric vehicles it arises without recovery only on very low-carbon grids. Maximizing the climate potential of this vast trade, the authors conclude, means extending circular-economy capacity across borders and matching vehicle technologies to the grids and recycling systems that will actually receive them.

Subject of Research: Life cycle assessment of the climate impacts of Japan's secondhand vehicle exports across powertrain types and destination countries

Article Title: Exporting sustainability? Assessing the environmental trade-offs of secondhand electrified and conventional vehicle flows from Japan to the global south and beyond

Article References: Yang, S., Teah, H. Y., Amasawa, E., Kanematsu, Y., & Kikuchi, Y. (2026). Exporting sustainability? Assessing the environmental trade-offs of secondhand electrified and conventional vehicle flows from Japan to the global south and beyond. Journal of Industrial Ecology. https://doi.org/10.1007/s44498-026-00193-y

Image Credits: AI Generated

DOI: 10.1007/s44498-026-00193-y

Keywords: used vehicles, electric vehicles, life cycle assessment, Japan, Global South, end-of-life recycling, grid carbon intensity, hybrids, emission offshoring, circular economy, transport decarbonization, aluminum recovery