How Do Financial Incentives Influence Electric Vehicle Charging Behavior?

The study investigates whether financial incentives can shift the timing of electric vehicle (EV) charging to off-peak hours. Using data from a field experiment in Calgary involving 150 EVs, researchers measured changes in charging behavior across three groups: one with financial rewards, one with informational prompts, and a control group. Financial incentives proved highly effective, reducing peak-hour charging by 49% and increasing off-peak charging by 37%. The informational prompts alone had no impact. When financial incentives ended, charging behaviors reverted, highlighting that financial motivation is key to shifting EV charging times and enhancing electric grid efficiency.

Central Research Question

The primary research question addressed by this study is: Can financial incentives influence electric vehicle (EV) owners to shift their charging behavior from peak to off-peak hours, thereby enhancing grid efficiency and reducing system costs?The researchers explore whether financial rewards are effective in encouraging EV owners to charge during low-demand times (10 PM - 6 AM) compared to informational prompts about societal benefits of off-peak charging, and if these behavioral changes are sustained once incentives are removed.

Previous Literature

The study builds on a growing body of literature on energy demand management and the potential for residential behavior shifts to stabilize energy grids. Studies have highlighted that EV adoption poses challenges for electricity systems due to increased demand, particularly if EV charging coincides with peak residential demand (5 PM - 8 PM). Prior research indicates that shifting energy consumption can alleviate stress on the grid, reduce greenhouse gas emissions, and lessen the need for costly infrastructure expansions. For instance, Burkhardt et al. (2023) found that homes with EVs in Texas responded well to overnight price discounts by shifting some of their electricity use. Similarly, Qiu et al. (2022) observed that households with EVs in Arizona responded to time-of-use rates, although customers voluntarily selected their preferred tariff in that study. Another study by Ito et al. (2018) found that financial incentives produced a more pronounced and sustained reduction in peak-hour energy use than non-financial motivations like social nudges. This research advances previous findings by providing empirical evidence on the specific effectiveness of financial incentives versus informational nudges in a real-world EV charging scenario, assessing both the degree of behavioral change and its durability after incentives are removed.

Data

The data was collected from a field experiment involving 150 EV owners in Calgary, Canada, conducted in partnership with the local utility ENMAX. EV owners were offered financial rewards and informational messages to see if these methods could change charging habits. The experiment was divided into two phases: a baseline period to establish initial charging behavior (February 2022 to March 2022), followed by two treatment phases. In Phase 1, participants were randomly assigned to one of three groups: (1) a “Rewards” group that received a financial incentive of 3.5 cents per kWh for charging during off-peak hours, (2) an “Info” group that received informational prompts on the benefits of off-peak charging, and (3) a control group with no intervention. Phase 2, implemented in September 2022, further divided the Rewards group into two subgroups: Rewards-Continue, which maintained the financial incentive, and Rewards-Stop, where the incentive was removed to evaluate whether the behavior change persisted without financial rewards. Charging behavior was tracked via in-vehicle monitoring devices, recording charge times, location (home vs. away), and energy use in kilowatt-hours (kWh).

Methods

The experiment employed a stratified randomization approach to assign participants to treatment and control groups, using a k-means clustering analysis to balance observable characteristics across groups. Charging behavior was monitored, and the study focused on within-day shifts from peak to off-peak hours rather than an overall reduction in electricity demand. The researchers analyzed data on two main variables: (1) whether a vehicle was charged during each hour (Charge Indicator) and (2) the kWh charged in each hour (Charge kWh). In Phase 1, researchers used linear probability and ordinary least squares regressions to assess the impact of financial and informational interventions, controlling for variables like temperature that could affect charging needs. Phase 2 followed a similar analytical approach but limited the sample to the Rewards-Continue and Rewards-Stop subgroups to determine the impact of discontinuing the financial incentive. Findings were also analyzed for heterogeneity in response by EV type (e.g., Tesla vs. non-Tesla) and driving frequency.

Findings/Size Effects

The experiment's results showed that financial incentives significantly influenced EV owners' charging habits, while informational prompts alone had minimal impact:

1. Phase 1 Results:

   • Rewards Group: Financial incentives led to a 49% reduction in peak-hour charging and a 37% increase in off-peak charging. On average, EVs in the Rewards group shifted 74% of their charging to off-peak hours, up from 56% during the pre-treatment period. These changes demonstrated a substantial degree of flexibility in EV charging compared to typical household electricity demand, which is often more rigid.

   • Info Group: Participants who only received informational prompts did not significantly alter their charging behavior relative to the control group, suggesting that moral or social nudges alone were insufficient to induce a change in charging habits.

   • Elasticity Estimates: The study calculated an own-price elasticity of -1.59 for off-peak EV charge timing and a cross-price elasticity of 2.10 for peak-hour charging relative to off-peak prices. These elasticity measures indicate that even modest financial incentives can lead to significant behavioral shifts within the same day, a finding that contrasts with the typically lower price responsiveness observed for household electricity consumption.

2. Phase 2 Results:

   • When financial incentives were removed for the Rewards-Stop group, peak-hour charging behavior reverted to pre-intervention levels, indicating the absence of habit formation. Participants in this group quickly resumed charging at peak times, while the Rewards-Continue group maintained their off-peak charging habits due to the continued incentive.

   • These results underscore the critical role of financial motivation in sustaining behavior change in EV charging habits, as the lack of long-term habit formation suggests that participants did not internalize off-peak charging as a norm but rather responded to the immediate economic benefit.

3. Robustness Checks:

   • The study included several robustness checks, such as excluding participants with solar panels (who may have unique incentives) and testing alternative temperature controls to confirm the stability of the findings. Results remained consistent, strengthening confidence in the main findings.

Conclusion

This study highlights that financial incentives, even if modest, are highly effective at shifting EV charging behavior from peak to off-peak hours, helping alleviate grid stress and reduce system costs. Unlike informational nudges, which showed no measurable effect, financial rewards significantly reduced peak-hour charging and increased off-peak charging. These findings have broad implications for policymakers and utility companies as they seek to manage rising electricity demand due to EV adoption. Effective demand-side management strategies that incorporate financial incentives could play a crucial role in managing EV-related load growth and integrating renewable energy sources by encouraging EV owners to shift charging times. The lack of habit formation observed when incentives were withdrawn implies that ongoing incentives or alternative approaches, like automated scheduling or dynamic pricing, may be necessary for sustained behavior change. Future research might explore other incentive structures or long-term impacts as EV adoption grows and grid flexibility needs become more pressing.

Citation

Bailey, M., Brown, D., Shaffer, B., & Wolak, F. (2024). "Show Me the Money: A Field Experiment on Electric Vehicle Charge Timing." American Economic Journal: Economic Policy. DOI: 10.1257/pol.20240250​.