Should Electric Vehicle Purchase Subsidies Be Linked With Scrappage Requirements?

This study examines whether linking electric vehicle (EV) purchase subsidies with scrappage requirements for old gasoline vehicles improves environmental and market outcomes. Using a simulation model calibrated with microdata from new vehicle transactions and used vehicle disposals, the authors find that the linked policy reduces new EV sales and carbon emissions less effectively than an unlinked subsidy. Linking reduces purchase additionality due to opportunity costs and eligibility limits, but increases scrappage additionality. Emissions reductions are smaller and costlier per ton under the linked policy. Approximately 75% of households receive no benefit under the linked subsidy due to these constraints.

Full Citation and Link to Article

Kevin Ankney & Benjamin Leard. (2025). Should electric vehicle purchase subsidies be linked with scrappage requirements? Journal of Policy Analysis and Management, 44(2), 553–578. https://doi.org/10.1002/pam.22639 

Extended Summary

Central Research Question

The central question explored in this study is: Should electric vehicle (EV) purchase subsidies be linked with scrappage requirements for used gasoline vehicles? Specifically, the authors examine whether such a policy design accelerates fleet turnover, improves emissions outcomes, and increases the cost-effectiveness of subsidies relative to unlinked EV subsidies. The paper assesses trade-offs between increasing EV sales and scrapping older, more polluting vehicles, with particular attention to policy efficiency and distributional equity.

Previous Literature

This study contributes to three key strands of literature. First, it builds on research evaluating vehicle scrappage programs, such as the U.S. “Cash for Clunkers,” which aimed to increase sales of fuel-efficient vehicles by offering purchase incentives tied to scrappage. While prior research (e.g., Li et al., 2013) found the program increased sales and reduced CO₂ emissions, it did so at high cost and with substantial non-additional behavior. Second, the paper engages with literature on subsidy design and “additionality” in environmental policy, a concept emphasizing whether subsidized behavior would have occurred without the incentive. Studies like Bento et al. (2015) and Boomhower and Davis (2014) highlight the importance of reducing non-additional payments to improve program efficiency. Third, the study engages with vehicle market modeling literature that estimates elasticities of demand and price responses in the new and used vehicle markets.

Data

The analysis draws from a rich dataset including:

• New Vehicle Customer Survey (NVCS): A nationally representative survey administered by Ipsos in 2018 with approximately 250,000 respondents, representing 1–2% of new vehicle transactions. It includes detailed information on new vehicle purchases, disposal methods, trade-in values, vehicle age, and household demographics.

• Consumer Expenditure Survey (CEX): Used to characterize household vehicle holdings and calibrate the distribution of used vehicles in the fleet.

• EPA and Argonne National Laboratory Data: Used to estimate vehicle lifetime emissions and fuel economy.

• Vehicle Market Data: Manufacturer suggested retail prices, lease rates, and EV eligibility for tax credits are drawn from Cox Automotive, InsideEVs, and cars.com.

• Supplementary economic indices: Including the FRED used vehicle price index and price elasticity estimates from prior literature.

These sources enable the authors to construct a national simulation model that mirrors real-world decision-making by households regarding new vehicle purchases and vehicle disposals.

Methods

The authors develop a two-part simulation model consisting of:

1. Vehicle Purchase Model: A discrete choice model that estimates household decisions to purchase an EV or a gasoline vehicle. Households are divided into two groups—buyers and lessees—to reflect differences in income, price sensitivity, and eligibility under the Inflation Reduction Act (IRA). The model incorporates utility maximization, where the net subsidy value is affected by vehicle eligibility, trade-in value, and vehicle age.

2. Used Vehicle Scrappage Model: An equilibrium model that simulates changes in scrappage behavior under a linked subsidy. The model accounts for “scrappage additionality”—scrapping that would not have occurred without the policy—and includes adjustments for equilibrium market effects. The authors incorporate rebound effects where the scrappage of vehicles may raise used vehicle prices, thereby lowering scrappage rates for remaining vehicles (a “Gruenspecht effect”).

The policy scenarios compared include:

• Unlinked Subsidy: As enacted under the IRA, offering up to $7,500 for new EV purchases without requiring scrappage.

• Linked Subsidy: A proposed model (based on Senator Schumer’s 2019 Clean Cars for America plan) that requires scrapping a gasoline vehicle at least 8 years old to qualify for the same subsidy.

Metrics evaluated include EV sales, scrappage rates, subsidy cost per ton of CO₂ reduced, and income distribution of benefits. Additionality is calculated for both purchase and scrappage behavior.

Findings/Size Effects

The study reports six major findings:

1. Lower EV Sales under Linked Policy: Linking a subsidy with a scrappage requirement reduces new EV sales relative to the unlinked IRA-style policy. This decline is attributed to participation constraints: only households with eligible gasoline vehicles (at least 8 years old and with a trade-in value below the subsidy) qualify. Approximately 75% of households are excluded due to these criteria.

2. Reduced Purchase Additionality: The effective value of the subsidy under the linked policy is lower for many households because the trade-in vehicle’s value often exceeds the subsidy. This “opportunity cost” diminishes the net benefit of participation, reducing the incentive to alter behavior. As a result, fewer EV purchases are attributable to the policy.

3. Higher Scrappage Additionality: Despite lower EV uptake, the linked policy leads to significantly more scrappage that is additional. Most participants in the linked policy would have otherwise traded in or sold their used vehicles, not scrapped them. Thus, the policy succeeds in removing older gasoline vehicles from the fleet.

4. Offsetting Market Effects (Leakage): The model estimates that about 20% of the scrappage gains from the linked policy are offset due to equilibrium price adjustments in the used vehicle market. As supply decreases, used vehicle prices rise, delaying scrappage of other vehicles and partially negating the intended environmental impact.

5. Lower Emissions Reduction Efficiency: The linked policy achieves fewer CO₂ reductions at a higher cost per ton compared to the unlinked policy. While the scrappage component helps retire higher-emitting vehicles, this effect is modest, and the high participation constraints lower the overall emissions benefit.

6. Greater Progressivity: The linked subsidy is more progressive than the unlinked policy. Lower-income households tend to own older, lower-value vehicles that are more likely to qualify for scrappage. High-income households are disproportionately excluded because they tend to own newer vehicles or vehicles with trade-in values above the subsidy amount.

Quantitatively, the average linked policy subsidy value (accounting for eligibility and trade-in opportunity cost) is about $1,000—far below the nominal $7,500. Only about 25% of households derive any benefit from the policy. The unlinked subsidy provides higher purchase additionality and greater overall emissions reductions per dollar spent.

Conclusion

Linking EV purchase subsidies to gasoline vehicle scrappage requirements introduces significant trade-offs. While it enhances scrappage additionality and improves the equity of benefit distribution across income groups, it substantially reduces participation and emissions reductions due to higher eligibility thresholds and opportunity costs. The authors find that unlinked subsidies, such as those provided under the IRA, are more effective at incentivizing additional EV sales and reducing CO₂ emissions. However, linked subsidies may be valuable when the policy goal is to accelerate the retirement of older, polluting vehicles, particularly among low-income households.

Ultimately, the study suggests that pursuing both objectives—boosting EV adoption and increasing scrappage—may require separate or hybrid policy instruments rather than a single linked mechanism. For policymakers, this analysis offers evidence-based guidance on designing cost-effective and equitable vehicle subsidy programs that support climate and transportation policy goals.