EV Incidence and Political Party Registrations by City

As this is being written on Election Day, and with an enormous chasm between the environmental/climate change plans of the two presidential candidates, we thought it a good time to look at how EVs fit into the CT political landscape based on voter registration in Connecticut.

I do not have data at the individual person level. I am working with counts at the city level – of EVs, median household income, and voter registration.

The most highly-correlated factor with respect to EVs is income. With Tesla the dominant, and pricey, make, EVs still carrying a higher cost than ICE generally, and limited supply of affordable used EVs, that isn’t surprising. Also, there is still a significant lack of access to charging in our more urban areas with many people living in multiple unit dwellings. For that reason, in the charts below, I have filtered out the roughly one-third of cities with a median household income of <$75K in order to obtain a sharper focus on the political registrations.

The chart at the top shows EV incidence and voter registration counts by party by city. The bars are all the same size because they total back to 100%. The variations in proportions by each color are driven by the proportion of voter registrations, which come from CT.gov, by party. Red and blue are obvious. The gold represents both independents and minor party registrations. Minor parties are a very small part of that grouping. The line shows EVs as a percentage of all vehicles within each city.

This screengrab is an excerpt. The full chart has been added to the EV Dashboard.

Bipartisan presence of EVs

There is not a significant correlation between voter registration profile and EV incidence. The two top EV cities, Westport and Weston, are Democratic redoubts. The next two cities are New Canaan, where registered Republicans outnumber Democrats by 2:1, and Greenwich which also has a Republican skew. This is followed by Wilton, which is evenly divided, and then Darien, which has a similar profile to New Canaan. We regard this bipartisan profile to be encouraging and feel that this, and the environment in general, shouldn’t be a partisan issue.

2016 Election Profile

CT is a blue state and voted for Clinton in 2016. The two charts below filter the chart by which cities voted for Clinton vs. Trump.  The hypothesis was that the profiles might be more extreme than overall registration, and that turns out to be the case. (The income filter remains in these two charts.)

Looking at the data this way, as presumably, Trump won the most conservative cities, a clear difference emerges with much higher EV incidence in Clinton cities, including Republican-dominant cities that voted for Clinton.

Clinton Cities 2016

Trump Cities 2016

Looking at the data this way, as presumably, Trump won the most conservative cities, a clear difference emerges with much higher EV incidence in Clinton cities, including Republican-dominant cities that voted for Clinton.

Spike in Model 3 Rebates leads to Slightly Stronger Rebate Activity in September

The September data were published on Friday, Oct. 30th, and show 84 rebates awarded with a $104,000 spend. Also, August was restated with rebates increasing from 40 to 44. A restatement of the prior month is common with these data releases.

The base-level trim of the Model 3 can still qualify for a rebate, even under the lowered $42,000 MSRP cap, and when those numbers are up, it raises the overall level. There were 37 Model 3 rebates, followed by 15 from the Chevy Bolt, possibly driven by some significant discounting. The spend level was $104,000, still pacing well under the allocated budget.

Rebate awards total $402,000 for all of 2020 through September against an annual budget of $3,000,000 (less admin and dealer incentives).

Pending Vote

CHEAPR changed the size of the rebates and the MSRP eligibility cap in 2019, which led to a large drop in the number of rebates awarded and the dollar amount spent. This was done at the time out of concern for the possibility of funds running dry late last year. Ever since then, there has been an announcement on the CHEAPR home page that revised rules will be coming in 2020. New rules were finally proposed in July. There was much disagreement about the proposal. Subsequent meetings in August, September, and October failed to resolve differences. No proposal has yet to be brought up for a vote. No meeting date is posted as of Nov 1. The CT EV Coalition does not like the incentive structure as originally proposed.

DEEP has asked their consultant, the CSE to go back and model additional scenarios. There are a number of variables in play, including an income-limited used EV incentive, an income-limited supplemental incentive, temporary stimulus incentive during this period of a weak economy, size of the rebate, and MSRP budget cap. We have blogged about a number of these issues before – here and here most recently.

The biggest sticking point, in my opinion, is the MSRP cap. At $42K, it is lower than neighboring states – NJ ($55K), MA ($50K), NY ($60K). More to the point, there just aren’t many BEVs that qualify. Below is the count of rebates by BEV model for 2020 to date.

There are only 7 models receiving rebates and just 4 that received more than single digits. If we exclude the Model 3 as our estimates are that ~75-80% of them are not eligible, and the eGolf, which is being discontinued, that leaves only 5 that are eligible, 3 with more than single digits. The eGolf is being replaced with the ID4, which will be ineligible. A loaded Bolt or Leaf Plus will exceed the threshold. The new Ford Mach-E begins at $43K. And, of course, the base trim level of the Tesla Model Y is over $42K. We feel CHEAPR needs to support the new generation of EVs, which include popular SUV or crossover form factors. Let consumer choice dictate where the rebates go and not put a thumb on the scale.

This will be the last CHEAPR post prior to the CHEAPR board meeting on Thursday, September 10.

Stats Update

The July stats have been published and rebate levels increased slightly over the desultory levels where they have been. There were 57 rebates in July, up from 46 in June. The numbers last year were 179 and 142 for June and July, respectively.

CHEAPR has spent $362,500 through July, plus another $40K or so on dealership incentives, out of an annual budget of $3 million.

9/10 Board Meeting

The published agenda does not include a vote. At least that’s what it says. Some key points:

• Despite DEEP’s not soliciting public comments on the MSRP cap and base rebate levels, many spoke up about them. The CSE was asked to scenario model and are expected to present their work. It is hard to think of a more difficult modeling environment than the present. The big question, of course, is that while the program has underperformed ever since the levels were changed in October 2019, there is an unknown with respect to the take rate for the supplemental LMI and used EV incentives that are likely to be adopted.
• DEEP’s position was that e-bikes cannot be statutorily defined as vehicles for the purposes of inclusion in CHEAPR. However, there is an agenda item about e-bike rebates.
• During the July meeting, there was a gap of roughly $800K between funds spent on rebates and available funding. A more detailed report on the CHEAPR budget is due. If any preliminary information has been released, we have not seen it.

Whenever we have looked at CHEAPR, it has appeared to be a worthwhile program. (Our complaint is with how the parameters were changed in October 2019.)

Creating a Comparison of Registered Vehicles with CHEAPR Rebates

For this analysis, which is seen in the chart at the top of the post, I began with all of the vehicles that were new to the file (comparing the January 2020 and July 2020 files). I then filtered that to the definition we have been using for new (as opposed to pre-owned) vehicles, which is the model year of 2019 or later. From that sub-set, I further winnowed it down to vehicles eligible for a rebate by make. Finally, I lined up the CHEAPR rebate data for the corresponding time period (first half of 2020), also by make.

There is some estimating here and not only for identifying new vehicles. Mostly, I used make as my proxy. For example, all new Chevrolets deemed new are considered eligible. This would only apply to one model, the Bolt, and it is possible to get a fully loaded Bolt that would exceed the MSRP cap and thus be ineligible. The same is true for the Mitsibushi Outlander. In the case of Tesla, I excluded all of the Model S, X, and Y vehicles, and took 25% of the Model 3. It is pretty easy to get above $42,000 with Model 3 options (e.g. long-range, all-wheel drive), so I made the assumption that relatively few are eligible.

If one takes these leaps of faith with me, then it works out that 53% of eligible vehicles are associated with a CHEAPR rebate. Of course, the CHEAPR restriction of only one rebate means that only first-time EV buyers can qualify. So it seems like the indicia are still pretty strong with respect to CHEAPR “driving” EV sales. If the rebate and MSRP cap were higher, CHEAPR could drive more EV sales.

If the chart at the top of the post is too dense to easily read, below is a chart with only the eligible and CHEAPR data points, and only the eligible makes.

Data from the Center for Sustainable Energy Helps Us Identify EV-friendly Dealers

We regularly field inquiries from club members and others asking for dealer recommendations. Usually, it follows a negative interaction with a dealership, when they walked in mistakenly thinking their inquiry about purchasing an EV would be well received. Not so fast!

It has been well documented, in the NY Times, in 2 Sierra Club shopper studies, and other reporting, that many dealers are indifferent or even hostile to EVs. But there are some dealerships that make an effort to sell EVs. To help guide consumers interested in non-Tesla EVs, we obtained from the Center for Sustainable Energy (CSE), the consultant that manages the CHEAPR incentive program for the Department of Energy and Environmental Protection (DEEP), the number of rebates by dealership from the program’s inception in 2015 through August 11, 2020.

I am using rebates as a rough proxy for sales/EV-friendliness. It’s the best we can do. You won’t find retailers of expensive vehicles, for example, a Jag or an Audi, on this list because the cost of the vehicles exceeds the MSRP eligibility cap. Consumers are eligible for one rebate lifetime, so repeat customers are not included. Some dealers may end up on our list in spite of themselves. But we can still use this directionally. Tesla is not included since it doesn’t have dealers.

We are covering a 5+ year period and understand that EV models come and go. Some manufacturers got out of the gate quickly (Tesla, GM, Nissan), while others came later to the party. The Chevy Volt, once the most widely registered EV in the state, has been discontinued. A couple of years ago, Honda introduced a PHEV Clarity that generated a fair number of sales. Since then, it has greatly slowed, reportedly due to distribution having been curtailed. There have also been 5 changes made during this period made by DEEP to rebate size and the MSRP price cap that determines eligibility. Finally, some dealers have multiple stores that were not separated in this dataset.

One-Third of Dealerships have not Awarded a Single Rebate

There are 270 franchised auto dealerships, according to their trade association (Connecticut Automotive Retailers Association) in CT. 185 of them have made a sale or lease associated with one or more rebates. Less than half, specifically 104, have disbursed 10 or more rebates and only 28, or about 10%, have awarded 50 or more rebates. (The denominator is somewhat inflated due to some dealers that don’t retail eligible plug-ins.)

The Top EV Dealers

These are the 5 dealers that have awarded more than 100 rebates.

• A-1 Toyota (New Haven)
• Honda of Westport (Westport)
• Richard Chevrolet (Cheshire)
• Karl Chevrolet (New Canaan)
• Lynch Toyota (Manchester)

Below are other top dealers for different makes that had between 50 and 100 rebates. Some makes haven’t had any dealer exceed 50 rebates.

GM – Ingersoll Auto (Danbury), O’Neill’s Chevrolet/Buick (Avon), H&L Chevrolet (Darien), Maritime Chevrolet (Fairfield), Grossman Chevrolet/Nissan (Old Saybrook), Chevrolet of Milford (Milford), Partyka Chevrolet (Hamden).

Toyota – Hoffman Toyota (West Simsbury), New Country Toyota of Westport (Westport), Middletown Toyota (Middletown), Hartford Toyota Superstore (Hartford), Westbrook Toyota (Westbrook)

Ford – Steven’s Ford (Milford), Stamford Ford/Lincoln (Stamford), Crowley Ford/Lincoln (Plainville)

Nissan – Grossman Chevrolet Nissan (Old Saybrook), Harte Nissan (West Haven), Crowley Nissan (Bristol)

BMW – BMW of Ridgefield (Ridgefield), BMW of Bridgeport (Bridgeport)

Finally, 2 stores that handle numerous brands:

Valenti Auto Sales (multiple locations) – Audi, VW, Porsche, Maserati, Fiat, Volvo, Alpha Romeo, Jaguar. (We presume most of the rebates come from VW.)

MJ Sullivan Automotive Corner (New London) – Chevrolet, Buick, Cadillac, Hyundai, Genesis

It should be acknowledged that this is a changing landscape. We are relying on the past as prologue to predict EV-friendliness and we hope it proves useful. As the EV landscape evolves and new models are introduced, we will update the data to the extent that it is available. We anticipate it will be. Going forward, the CSE has advised they will be making more granular data available with their normal releases of CHEAPR data.

The used car market, in general, is more than double that for new vehicles. That does not appear to be the case for EVs to this point.

CHEAPR Likely to Implement Used EV Incentive in 2021

CHEAPR, the CT state EV purchase incentive program, is considering offering incentives for purchases of used EVs. This incentive would be limited to lower and middle-income individuals/families. There are a number of changes being considered by CHEAPR, but with respect to used EVs, the legislature specifically authorized this incentive, the proposal was well received by the board, and the public comments were favorable. It seems a lock to happen, though there are below the line development tasks that will cause it to not be available until next year.

What is the State of the Used EV Market

We don’t have access to the data that would enable us to definitively answer this. But we have some information that may be useful for drawing inferences.

During the CHEAPR board meeting of July 17th, there was a presentation by the auto-dealership representatives on the board. They stated that there are few used EVs in the marketplace and the prices were low, creating an unvirtuous circle. They support the incentive and think that that it promises to sufficiently stimulate consumer demand so that dealers will be willing to bid more aggressively at auctions to augment the supply in the state.

The used EV incentive will differ from the new car incentive in that it will also apply to independent used car dealers. Used car dealers do not have to be affiliated with a manufacturer. A Google search for “used EVs for sale in CT” brought up a results page consisting of only independent dealers, mainly large ones like Carvana, Iseecars, and CarGurus. Those companies had both paid and organic listings on this first page of the search results. A search for “used Teslas for sale” brought up a largely similar set of sites, except that Tesla itself appeared, as it is in the business of retailing its own used vehicles. There is another company specializing in used Teslas called OnlyUsedTesla.com.

I suspect that the board members who represent the dealerships are not factoring Tesla into their thinking. For them, EVs are still a niche product and many of the non-Tesla EVs in the used marketplace are the first generation (read: low range) models. (We may be at a point where this is beginning to change as later model EVs are now coming off-lease.) And the dealers, based on the search results and their own words, aren’t making a serious effort to source and sell them. The fact that the independents are spending money on sponsored links indicates that there is at least a minimally viable business. Search is highly targeted and can yield a positive return on a small campaign.

Quantify Used EVs from the DMV File

To get some kind of quantification of used EVs relative to new, I went back to the file we recently got from the DMV of all registered EVs in the state as of July 1, just to get an idea of what was entering the market. My proxy for used EVs was vehicles added to the file between January and July with a model year earlier than 2019. This is a rough measure and is reflected in the chart at the top of the post. Each bar represents vehicles added to the file in the first half of the year sorted by make, with the orange portion being those that are categorized as used by our proxy measure. 22% of the EVs added to the file could be characterized as used based on this definition. 47% of the vehicles added are Tesla, but only 10% of those fit this definition of used.

• This, coupled with the information from the dealers, indicates a small used EV market at this point.
• Even though it is small, there is a used EV business.
• The fact that there is no franchise requirement begs the question of whether Tesla could sell used EVs in Milford (or elsewhere in the state) using the same rationale that led to their being able to lease. In the case of new vehicle leasing, customers still have to go out of state to pick up the vehicle. Would that be a requirement if they could sell used?
• The EV Club is supportive of a used CHEAPR EV incentive, but based on this information, along with the LMI restriction, we don’t expect that it will be disbursing large sums in 2021.
• It is important, as used and possibly other incentives, are incorporated into CHEAPR, that the stats page be updated to track them separately.

The CSE, DEEP’s consultant for CHEAPR, has been sent back to model new scenarios and we will see what they forecast.

Turnover Analysis of EV Makes Driving Adoption

When we build our semi-annual EV dashboard with data sourced from the DMV, we, of course, look at trends by vehicle make. The analysis in this post is intended to give a more focused look at the recent impact of the various EV makes by isolating the vehicles that departed between January and July 2020 and comparing that turnover to the new vehicles added in the most recent July file.

The chart above shows that Tesla has low turnover, coupled with a high number of new vehicles added. That is not a surprising result. When analyzing EV adoption data, the answer to almost every question is “Tesla.” But the impact is seen more starkly in this view compared to the “trends by make” or “waterfall” charts in the dashboard. Since the outsized presence of Tesla tends to overwhelm everything else, it makes it hard to visualize any movement that may exist elsewhere. The answer: show the data without Tesla.

What pops on this chart is the contribution increase from Hyundai. This may be an early signal of a serious EV push, followed this past week by the announcement that Hyundai is spinning off its Ioniq marque into a dedicated EV sub-brand (like Volvo with Polestar) and plans 3 new EV model introductions over the next several years, beginning with a mid-sized crossover in 2021.

Toyota, which has occupied a distant number 2 position over the past couple of years with its Prius Prime PHEV, showed a smaller increase on a lower base.

Ford is going in the opposite direction, with more EV turnover than additions. They have an eagerly anticipated launch in 2021 of the Mach-E, a crossover that bears the iconic Mustang logo.

Audi, Land Rover, and Subaru also spiked, but the numbers were low. Audi showed 9 departures and 34 adds for its new e-Tron. Land Rover, just entering the plug-in world (and separate from the Jaguar iPace), went from having 0 EVs to 16. Subaru had one departure and 19 adds.

37 CHEAPR rebates in June

This tepid number was only slightly higher than the 27 in May, closing the quarter with a soft 81 rebates total and 275 for the first half of the year. This 275 compares to 818 during the first half of 2019.

Partly, this was due to the recession, but a lot of it has to do with the changes made to the program in October 2019, when the price cap for vehicle eligibility was lowered from $50K to $42K. You can see in the graph that the numbers immediately tanked in November and have stayed low.

CSE Proposal for CHEAPR Program Revisions

The CHEAPR board met on July 17th to entertain proposed program changes submitted by their consultant, the Center for Sustainable Energy (CSE). These proposed changes were a decidedly mixed bag. DEEP is accepting public comments until August 12th. Email comments to the at deep.mobilesources@ct.gov

These are our positions:

Dealership Data Request

The club regularly gets requests for dealer recommendations. Usually, this follows a poor dealership experience. We have an informal list of some dealers that have been recommended by members, but it is not complete and doesn’t cover the whole state.

So we would like to let the data do the talking to the extent it can. Our objective is to publish a list of top dealers for each brand and we hope to have representation across the state.

We made the request to CHEAPR, which is run by the Center for Sustainable Energy (CSE), on June 23. On July 7, they advised that they were working on it. We are still waiting as of this writing on July 26th. Hopefully, we will receive it soon.

We will be using rebates is a proxy for sales and it is imperfect. When we publish the data, we will describe in detail what it represents and its limitations. We have asked for granular data. We will cut and sort the data to make it as clean as possible, perhaps by brand by county.

There was a recommendation in the DEEP EV Roadmap to create a dealer recognition program. We support this as long as it isn’t vaporware. If it proves to be truly robust and stays up to date, then we may back off of the data.

We hope to be able to post the actual data soon. This doesn’t seem like a difficult request.

EVs are not uniformly distributed across the state

Fairfield County has consistently tracked at around 40% of EVs in the state and is 41% in this July 1, 2020 iteration. This compares with its having 26% of the state’s population. All of the other counties under-index relative to population.

EV Distribution by City

The top cities largely held their position in terms of the number of registered EVs with Greenwich, Stamford, and Westport in the top 3 positions. This chart excerpt shows the most recent two data points, January and July, for the largest cities. Greenwich added the largest number of any city with 69 additional EVs since January.

EVs Per Capita by City

Westport remains the leader in EVs per capita, followed by Weston, New Canaan, Woodbridge, and Greenwich. This screenshot is an excerpt of the top cities. The full list can be viewed in the dashboard (use the scroller).