Today, a bipartisan group of legislators introduced a bill into the federal legislature that would strengthen the current EV tax credit in place in the United States. I’m not the biggest fan of it, and in this post, I’ll explain why.

My Values

To be clear, I’m no unbiased individual.

• I currently own shares of both $GM and $TSLA, both of whom are at a disadvantage under the current system of EV incentives.

• Keeping my bias in mind, I prefer to support companies that are first-to-market and pushing the boundaries of new technology. To the extent that technology subsidies are supposed to accelerate the adoption of that technology, I believe that supporting the early innovators and adopters is important.

• I like the idea of tying subsidies to performance, to the extent that it’s possible to do so and makes sense (i.e. there needs to be clear metrics on what defines “performance”, and it needs to be the case that those metrics are easy to measure).

• I’m also partial to the idea that, at least for this particular case, a subsidy passed by the US government ought to benefit US automakers.

• While EV subsidies are far from the worst way to spend money, I’d still prefer incentive structures where total payouts are lower than higher.

• I’d also prefer incentives to help make technology accessible for everyone, not just a small group of well-off individuals.

To be clear, I still believe that EVs are better products and that it would be optimal to accelerate their adoption.

The Current System

The current system is slightly confusing at first, but it isn’t too bad - the exact same system applies for both plug-in hybrids (PHEVs) and pure EVs (BEVs). Here’s a link to the IRS webpage that lays everything out.

Basically, each model has an associated credit:

• If the car is either a PHEV or a BEV, the subsidy is at least $2,500 (run-of-the-mill hybrids, like the Prius, get nothing).

• If the battery pack has a capacity of 5 kWh or more, add $417 to the subsidy.

• For each kWh of battery capacity over 5 kWh, add another $417, up until the subsidy hits $7,500.

• This implies that the max subsidy is hit when your battery stores at least 16 kWh of energy - the car is electrically driven, so there’s $2,500. The first 5 kWh give another $417, and the next 11 kWh give another 11*$417. This comes out to $7,504, which gets capped at $7,500.

Once manufacturers sell 200,000 PHEVs or BEVs in the US, the countdown timer begins:

• For example, Tesla sold its 200,000th EV in the US in July of 2018. This begins the countdown.

• For the calendar quarter in which the 200,000th car was sold and the quarter after that, nothing changes. In other words, since Tesla sold the vehicle #200,000 in July, then nothing changes in either the first quarter (July-September) or the one after that (October-December). Likewise, GM sold their 200,000th EV/PHEV in Q4 of 2018. This means that nothing changes for Q4 ‘18 or Q1 ‘19.

• For the next six months, the EV incentive is halved on PHEVs or EVs from that manufacturer. In other words, a car that previously would have earned $4,000 in subsidies now only earns $2,000. There’s no limit to the number of cars this can apply to. For Tesla, the subsidies were halved starting on 1/1/2019, GM’s EV subsidy phase-out began on 4/1/2019.

• For the next six months after that, the EV incentive is further halved. Again, there’s no limit on the number of sales. A car originally earning $4,000 in subsidies now earns only $1,000. This kicks in for Tesla on 7/1/2019, and for GM on 10/1/2019.

• Finally, after another six months, the manufacturer’s cars are ineligible for federal EV subsidies. Anyone buying an EV from Tesla after 1/1/2020 or a PHEV/EV from GM after 4/1/2020 gets nothing.

For clarification, the companies themselves don’t get any subsidy checks from the government. They do, however, enjoy more competitive post-subsidy prices. Here’s a helpful webpage that lists each car by manufacturer, how much of a subsidy it has, and when the subsidies begin to phase-out.

What today’s bill would do

Disclaimer: I haven’t read the bill myself, and am relying on the summary put out by Sen. Stabenow, along with articles put out by Bloomberg, The Hill, and Reuters. I don’t know the exact specifics.

What seems clear is that the structure doesn’t change too much. Essentially, after the first phase of 200,000 vehicles, another phase would kick-in, this time with capacity for 400,000 vehicles and a subsidy limit of $7,000 per vehicle. The phase-out period is reduced to around nine months, though I don’t understand how exactly that’s implemented (the 25% phase appears to be eliminated). There’s also a portion of the bill related to hydrogen fuel cell vehicles, but that’s not what I’d like to focus on.

While I’m no political expert, I don’t believe the bill to be likely to pass. I doubt President Trump wants to sign a bill supporting additional spending on EVs, and I don’t think a supermajority in Congress is likely, either.

Issues

Whether we analyze the potential bill or the current system, I believe that there are a few issues. Some are the result of bad design from the start, others are the victim of their own success - the number automakers either currently producing or planning to produce EVs is significantly higher than it was a decade earlier.

Tax credits aren’t the best incentive. For clarification, the subsidy is run as a tax credit received by the buyer. This creates several problems, the first of which is that you can’t use more of a tax credit than your tax liability, according to Edmunds. In other words, if you earn an EV tax credit of $7,500 but you owe $4,000 in taxes, you only get $4,000. You can’t roll the unused portion over to the next year, nor will you get a check for the difference. This structure, therefore, benefits higher-income buyers far more, instead of a system where the incentive is always the benefit that the buyer receives.

Tax credits also don’t grant a benefit at the time of purchase, and delayed gratification doesn’t work well for many end consumers. When purchasing a new vehicle, taxes and other fees can amount to thousands of dollars out of pocket at the time of purchase. The tax credit, however, must wait until the next April, creating a liquidity problem for buyers.

Ultimately, tax credits aren’t an optimal means of administering subsidies, and research agrees. Hardman et al. reviewed 35 papers on the influence of incentives on EV adoption and found that consumers valued incentives at the time of purchase more than incentives later on (e.g. sales tax exemptions given at the time of sale were more valuable to consumers than tax credits). Specifically, the studies found that sales and registration tax exemptions were more effective than other forms of incentives, e.g. toll waivers, free parking, and HOV lane access.

Incentives apply equally to expensive and mass-market vehicles. The same study-of-studies found that incentives are less important for more expensive vehicles (e.g. Tesla’s Model S or X, which today starts at $85,000 and $89,500, respectively). This is pretty intuitive, as $7,500 represents potentially less than 10% off of a base Model S but over 20% off of a base Model 3.

International sales make the current system easy to game. Note that the only sales that count toward the 200,000 vehicle limit are US sales. This means that hypothetically, a company could choose to reach 199,999 US sales at the end of the quarter, ship all other vehicles overseas, and then make its 200,000th sale on the first day of a quarter, lengthening the time for which incentives are available. Furthermore, companies abroad could grow their production abilities overseas, and only come to the US once their quarterly production rate is significant, allowing many more vehicles to qualify for the credit than, say, an American company that grows organically from the start. By the time both companies reach 200,000 US sales, it’s likely that the foreign company has a much faster sales and production rate, and thus benefits more from the subsidy.

Industry pioneers are held at a disadvantage. Consider the two companies which have already begun the incentive phase out - GM and Tesla. Tesla was an industry pioneer; GM was the first to sell a vehicle with usable range at a mass market price point. While these two companies face declining incentives, competitors such as Mercedes-Benz, Porsche, Hyundai, and Kia can all reap the benefit of the full subsidy. While Tesla and GM invested into R&D, battery prices around the world steadily dropped, and now other players are entering the market. Tesla and GM no longer have a competitive advantage for being first-movers in the market; if anything, the incentives have generated a competitive disadvantage. Back in 2009, such a situation may have been hard to predict; today, the market has shifted sufficiently.

Pardon my more nationalistic side, though I believe that American companies are being disadvantaged by American subsidies.

The current system stands to spend a lot of money. Suppose a manufacturer (say, Tesla) only built cars that qualified for the full $7,500 credit. Further suppose that they were able to sell 200,000 vehicles in the US, and that buyers generally were able to take full advantage of the tax credit. Even before accounting for future US sales during the phase-out, the subsidies have amounted to $1.5 billion. That number is complicated by many factors: as I’ve already mentioned, it’s unclear how many additional vehicles they’ll sell in the US and at what subsidy level and it’s unclear whether US buyers took full advantage of the tax credit. For other manufacturers, it’s unclear if Porsche will sell 200,000 EVs by the time the current system has disappeared. Nevertheless, here’s a list of some manufacturers: BMW, Ford, GM, Honda, Hyundai, Jaguar, Kia, Mercedes-Benz, Nissan, Porsche, Tesla, Toyota, Volkswagen, and Volvo. Each one has sold vehicles that had some level of subsidy attached to them (not all vehicles had the full $7,500; some were PHEVs). Nevertheless, if those manufacturers alone averaged out to $1.5 billion each, the total bill would come to $21 billion. [1] Furthermore, there’s no limit to the amount that we might spend. New startups or existing automakers that I haven’t listed could explode into the EV market and sell more cars.

Superior products enjoy similar benefits as inferior ones. Ultimately, the largest drawback that the current system presents is that it subsidizes barebones EVs just as much as full-fledged ones. As an example, suppose that automaker A comes to market with an EV that gets a range of 204 miles (according to the EPA), while automaker T sells an EV with 295 miles of range (again, according to the EPA). Holding everything else constant, should we subsidize the two vehicles equally? Going further, suppose that company B makes an EV can travel 114 miles on a charge, yet has a battery pack larger than 16 kWh. Does it make sense to subsidize company B’s product as much as company T’s? If one of the goals of electric vehicle incentives is to support electric transportation, and vehicles with lower ranges get used less than vehicles with longer ranges [2], then surely we ought to align incentives with the number of electric miles that the average person travels.

A few small tweaks

Here’s a small, basic improvement that I believe would fix the issues of total spend and pioneer disadvantage: decide the total amount of money to be spent, and then hand out subsidies until the total pool is exhausted. For example, if we decide that we’d like to spend a total of $10 billion for subsidizing EVs, create a $10 billion pool and allocate it out to the various manufacturers as they sell cars. Not only would total spend be limited, but the incentive structure is now set up to benefit whichever company produces the most EVs. Companies are no longer penalized for selling too many EVs while their competition catches up.

Of course, this could be combined with several other smaller proposals. Perhaps we deliver the subsidy at the time of purchase as a sales tax reduction/exemption, or allow the income tax credit to be rolled over to future years. Perhaps we introduce a maximum price for which vehicles qualify. Perhaps we restructure the incentive calculation to account for the technological progress made over the past decade.

a more encompassing re-do

All of the previous proposals combine into what I’d propose as a new system. I call it the “quarter roll-down”. Here are my proposals:

• Subsidies should be based on the EPA tested electric-only range. This allows both PHEVs and EVs to use the same incentive system without too much confusion. I’m not 100% familiar with the method that the EPA uses to obtain range ratings for EVs, so changes might need to be made here. However, this finally allows us to treat better products, better.

• In addition to a total incentive pool size (say, $25 billion), introduce a “sunset date” of say, 5 years from the date of passage of the bill. Once the total incentive pool has been depleted, or the sunset date has passed, the incentive program is over.

• For each vehicle, calculate the incentive as miles of electric-only EPA range x quarters until sunset date. In other words, an EV with 300 miles of range sold at the beginning of the program (i.e. 20 quarters, or 5 years, remaining) would receive a subsidy of $300 x 20 = $6,000. With only 3 years remaining, the same vehicle would receive $300 x 12 = $3,600. This allows incentives to diminish over time, as technology improves and battery prices drop.

• Remove all manufacturer limits on incentives. If a car is sold in the US while the program is ongoing, it gets the incentive.

• Restructure incentive payouts as an immediate sales tax deduction. If there’s excess credit, allow it to be used as an income tax payment, and allow excess tax paid to be refunded.

• Vehicles with a sticker price of $50,000 or higher are unable to receive subsidies.

I’d love to hear about what y’all think. I’m sure that there are a few points that I’m missing - please let me know. Also, a lot of the information I found is based off of a term paper I wrote last semester. If you’d like to read, you can find a copy here.

pushing back

Should EV incentives really be limited to certain price brackets? A very valid point is that more expensive cars tend to get worse mileage; replacing those cars with EVs is better for the environment. Nevertheless, something feels wrong to me about subsidizing more expensive vehicles. I’m certainly willing to concede that I’m being illogical, but consider a hypothetical electric sports sedan that reaches a price of $130,000. Would this buyer really need an incentive in the thousands to convince him/herself to buy an electric vehicle? Ultimately, however, this is not the hill I’d die on.

[1] It’s unclear, to me at least, whether different brands under the same parent company would individually qualify for the EV incentives.

[2] The California Air Resources Board published a 2017 study calculating the number of electric miles driven by various PHEVs and EVs. Model S drivers averaged around 13,500 miles each year, while Volts averaged around 7,500 electric miles each year. Meanwhile, shorter-electric-range PHEVs like the Ford Fusion Energi traveled only about 2,500 electric miles each year, on average.