THE RESHUFFLE OF ELECTROMOBILITY

THE INHERITED INDUSTRY

Customers buy usable journeys

An electric car must do three things. It must complete the journeys its owner needs, reach dependable charging and be affordable to buy or lease. A failure in any one can stop the sale.

THE INHERITED INDUSTRY

Carmakers controlled the vehicle, not the fuel network

Carmakers specified the finished vehicle, integrated supplied parts and remained responsible when the vehicle failed. Dealers handled the customer. Fuel supply sat outside this structure: carmakers did not build or operate petrol stations.

THE INHERITED INDUSTRY ·US passenger ·2008-2012

Cell cost blocked an affordable electric car

From 2008 to 2012, battery cells cost too much for a mass-market electric car. Carmakers could simplify the powertrain, but they could not reach a price ordinary buyers would pay. Cell cost was the main constraint.

THE INHERITED INDUSTRY ·US passenger ·2008-2012

Rated range does not guarantee a usable journey

Rated range is only a test result. Buyers care about real-world distance, charging on longer trips and whether they trust those chargers to work. A car can handle every ordinary day and still fail the occasional journey that decides the purchase.

THE INHERITED INDUSTRY ·US passenger ·2008-2012

The vehicle changed before the industry structure did

Early electric cars proved that the technology could work. Carmakers still integrated the vehicle, suppliers still served them and the carmaker remained responsible for the finished product. The next six chapters show where that inherited structure changed and where it did not.

RESHUFFLE 01 - THE SYSTEM ·US passenger ·2012-2016

The vehicle worked, but some journeys did not

Tesla showed that an electric car could serve normal daily travel. A household could still reject it because one necessary long trip depended on public charging. Battery cost remained relevant, but route coverage became the main constraint for this customer group.

RESHUFFLE 01 - THE SYSTEM ·US passenger ·2012-2016

No established industry owned public charging

Carmakers had never supplied fuel. Utilities delivered electricity to a meter, property owners controlled parking and public authorities controlled streets. None was responsible for making electric journeys possible across homes, cities and motorways.

RESHUFFLE 01 - THE SYSTEM ·US passenger ·2012-2016

Five types of organisation built charging

Governments funded coverage. Electrify America arose from Volkswagen's diesel-emissions settlement. Eight carmakers formed a consortium. Independent operators pursued charging revenue. Tesla built a network for its own customers. In Norway from 2010 to 2015, public-charger subsidies produced more than twice as many electric-car sales per krone as vehicle subsidies.

Scope: the Norwegian comparison covers chargers open to all drivers. It excludes Tesla's proprietary network.

RESHUFFLE 01 - THE SYSTEM ·US passenger ·2012-2016

Charger count did not measure reliability

A charger must exist in the right place, work when the driver arrives and report accurate live status. Today about a third of stations on the interstate fast-charging network publish live status, and the two largest operators publish none. A 2025 model estimated a 1.1-point increase in the electric share of 2030 US sales from live information alone. Information combined with reliable chargers raised the estimate by 8.0 points.

Scope: the station share is measured on six interstate corridors. The 1.1- and 8.0-point results are forecasts for 2030, not observed sales.

RESHUFFLE 01 - THE SYSTEM ·US passenger ·2012-2016

Tesla reduced the constraint for its own customers

Tesla built and operated a charging network integrated with its vehicles and route planning. This made long-distance travel easier for Tesla drivers. Other drivers continued to depend on public networks, and households without private parking still faced a different access problem.

• Control, no measured capture
• Value moved

Scope: Tesla reduced the constraint for its own US premium customers, not for the wider market. Tesla does not report Supercharger profit separately.

RESHUFFLE 02 - THE ARCHITECTURE ·US passenger ·2017-2020

The predicted supplier takeover did not occur

Bain predicted in 1999 that large suppliers would design standard systems across carmakers and capture more industry profit. The prediction followed the computer industry. Jacobides and MacDuffie found a different automotive result. Carmakers maintained a relatively stable share of industry market value through 2005, while computer makers' share fell sharply.

Scope: the market-value comparison ends in 2005. It cannot by itself explain what happened during the electric-vehicle transition.

RESHUFFLE 02 - THE ARCHITECTURE ·US passenger ·2017-2020

Producers moved the other way

Instead of separating, the industry integrated. Researchers examining battery-pack sourcing found every serious producer, established and new, moving toward vertical integration or close alliances with a cell maker over the same period. No contract manufacturer of electric vehicles emerged. Tesla integrated further with each decision on products, production and supply; BYD, which became a carmaker by buying one, was at least as integrated.

Scope: this is what MacDuffie, Jacobides and Tae report from battery-pack sourcing. It describes what producers did, not why they did it.

RESHUFFLE 02 - THE ARCHITECTURE ·US passenger ·2017-2020

Simpler propulsion did not remove integration work

An electric powertrain has fewer mechanical parts than an engine powertrain. The vehicle still requires hardware, software and safety systems to work together. A decade into the transition, no electric-vehicle contract manufacturer had taken over the carmaker's integration role. Carmakers also brought important software work back into internal teams.

RESHUFFLE 02 - THE ARCHITECTURE ·US passenger ·2017-2020

Carmakers invested without one proven priority

US electric-car sales faced three constraints at once: expensive cells, the cost of dedicated platforms and uneven charging access. No reliable comparison shows which mattered most. Carmakers had to commit billions without knowing whether any single investment would unlock demand.

RESHUFFLE 02 - THE ARCHITECTURE ·US passenger ·2017-2020

Companies shared platforms, not open modules

Carmakers had long resisted standard components sold unchanged across rival vehicle makers. Reuse later moved to larger integrated platforms. Rivian and Volkswagen announced a jointly developed platform, while CATL offered a supplier-owned chassis. Each platform remained controlled by its owner or joint venture.

Scope: both platform examples are announced plans. Their commercial results are not yet available.

RESHUFFLE 02 - THE ARCHITECTURE ·US passenger ·2017-2020

Carmakers controlled changes to vehicle behaviour

Software can change vehicle functions and battery management after sale. The carmaker controls validation, the update channel and access to the vehicle's internal network. A software company can supply applications or dashboard services without gaining authority to release safety-critical changes.

• Revenue disclosed, cost not

Scope: one carmaker reports software and connected-service revenue inside a four-part bundle. No carmaker reports software costs separately. The figures cannot show software profit.

RESHUFFLE 02 - THE ARCHITECTURE ·US passenger ·2017-2020

Software and vehicles operate on different clocks

Software teams can release changes in weeks. Vehicle hardware and safety approval operate over years, and the car must remain safe for much longer. Carmakers and software firms therefore partnered rather than one replacing the other. The carmaker remained legally responsible for the finished vehicle and retained authority over safety-critical integration.

• Control, no measured capture

Scope: carmakers control vehicle operation and software release. The wider customer relationship remains contested, and no cited carmaker reports software profit separately.

RESHUFFLE 03 - THE SUPPLY BASE ·US passenger ·2020/21-2023

The battery returned as a supply constraint

Battery cells constrained the industry twice for different reasons. First, high cost made vehicles too expensive. During the early-2020s shortage, carmakers could not obtain enough cells for planned production. Engineering and scale address cost. Long-term supply commitments address allocation.

RESHUFFLE 03 - THE SUPPLY BASE ·US passenger ·2020/21-2023

Carmakers paid to secure cell volume

Carmakers signed take-or-pay agreements, prepaid suppliers, bought equity stakes and formed manufacturing joint ventures. These commitments reserved supply and transferred demand risk to the carmaker. They were responses to allocation, not ordinary price negotiation.

RESHUFFLE 03 - THE SUPPLY BASE ·US passenger ·2020/21-2023

CATL earned profit without taking vehicle control

CATL reported first-half 2024 revenue of CNY 166.77 billion and profit of CNY 22.87 billion. Its gross margin was 26.5%, and it supplied 37.5% of global electric-car batteries. Carmakers still specified the pack, integrated the vehicle, carried the warranty and remained responsible to customers and regulators.

• Measured outcome
• Value moved

Scope: CATL is one audited company. Its profit and Ford's losses are separate results; they do not prove that money moved directly from one to the other.

RESHUFFLE 03 - THE SUPPLY BASE ·China passenger ·2020/21-2026

In China, eligibility rather than scarcity constrained battery access

China had abundant cell production. A carmaker could still lose market access if its battery did not satisfy localisation and supplier requirements. The constrained object was the battery, but the decision sat with the regulator.

RESHUFFLE 04 - THE RULES ·China passenger ·2020/21-2026

Battery capability and battery eligibility were different

China had extensive cell manufacturing, capacity and technical capability. Selling a vehicle still depended on where its battery came from and whether its suppliers qualified. A mature domestic industry could therefore remain part of the market-access constraint.

RESHUFFLE 04 - THE RULES ·China passenger ·2020/21-2026

The battery failed the test, but the rule caused the failure

The vehicle became ineligible because of its battery. A cell maker could not change the eligibility rule. The regulator that set the rule controlled the decision, while the carmaker could respond by changing suppliers, contracts or production location.

RESHUFFLE 04 - THE RULES ·China passenger ·2020/21-2026

Rules can change demand, supply or finance

A purchase incentive lowers the buyer's price. A supplier rule decides which components a carmaker can use. A financing rule changes the monthly payment. Vehicle sales alone cannot show how each rule changed the industry.

Scope: the supply example comes from China and the financing example from the United States. The comparison is ours; no single study proposes this classification.

RESHUFFLE 04 - THE RULES ·China passenger ·2020/21-2026

China identified critical technologies in 2001

In 2001, the Chinese state identified batteries, motors and control systems as domestic priorities. Later policies linked vehicle eligibility to components and approved suppliers. Carmakers changed suppliers, contracts and investment to remain eligible.

• Control, no measured capture
• Value moved

RESHUFFLE 04 - THE RULES ·China passenger ·2020/21-2026

China's charging system is visible; its effect on sales is not

About half of Chinese electric-car owners have a home charger. Another third use shared chargers in apartment parking, although only 7% live in single-family houses. These figures describe access, not lost sales. We also have too little evidence to judge whether grid capacity constrained the Chinese market.

Scope: charging provision in China is documented, but its effect on sales and its financial returns are not. The sources say even less about grid constraints.

RESHUFFLE 05 - THE TRANSACTION ·US passenger ·2024-2026

US electric sales weakened without a technical setback

By 2024, electric vehicles could still serve the same journeys as before. Battery capability had not suddenly declined. US sales nevertheless weakened sharply because the transaction joining vehicle, buyer, lessor, lender and government support became less attractive.

Scope: no single figure isolates the cause of the sales decline. Purchases brought forward before the tax-credit deadline also affected its timing.

RESHUFFLE 05 - THE TRANSACTION ·US passenger ·2024-2026

Lease residual assumptions proved too high

Leases written in 2022 and 2023 assumed vehicles would retain about 50% of their value. Reported resale values were nearer 35% to 40%. More than 300,000 returns were expected in 2026, compared with 123,000 in 2025. Those vehicles return to a market where most buyers purchase second-hand. Hybrid sales rose while battery-electric sales fell, showing that the weakness was specific to the battery-electric offer.

Scope: the return volumes are projections from a dealer-software company. The hybrid comparison identifies a difference in demand; it does not explain why buyers made that choice.

RESHUFFLE 05 - THE TRANSACTION ·US passenger ·2024-2026

Federal credits helped build the leasing channel and then ended

A leased vehicle counted as commercial, which put it outside the sourcing requirements that applied to a purchase. The credit therefore selected a way of buying rather than a price. Leasing rose from about 15% to roughly two-thirds of US electric-car sales. Public Law 119-21 then ended the consumer, used-vehicle and commercial credits for vehicles acquired after 30 September 2025.

Scope: the charging-equipment credit ended later, on 30 June 2026. Purchases brought forward also contributed to the fall after the vehicle-credit deadline.

RESHUFFLE 05 - THE TRANSACTION ·US passenger ·2024-2026

Ford Model e reported three years of large losses

Ford Model e reported losses of $4.7 billion in 2023, $5.1 billion in 2024 and $4.8 billion in 2025. This is one of the few directly measured economic outcomes in the publication.

• Measured outcome
• Value moved

Scope: Model e includes company-wide software and connected-vehicle costs that Ford does not separate. It is one US company, not the whole industry.

RESHUFFLE 06 - THE POWER SYSTEM ·Depot fleets ·2020-2026

A fleet's operating job changed the constraint

A delivery fleet runs known routes and returns many vehicles to one site each night. More electric trucks were sold in the United States in one year than in the previous seven combined. Vehicle availability improved, but simultaneous depot charging concentrated electricity demand at one location.

RESHUFFLE 06 - THE POWER SYSTEM ·Depot fleets ·2020-2026

The route was feasible, but the depot lacked power

The truck could complete its assigned route. Connecting enough power to the depot could take nine months to three years, or five years when the local network required an upgrade. The utility connection, not vehicle range, controlled the launch date. No one could sell a fleet a place further up the queue: California law bars a utility from discriminating between customers, and in 2024 the regulator refused to prioritise electric-vehicle projects for that reason.

RESHUFFLE 06 - THE POWER SYSTEM ·Depot fleets ·2020-2026

Some fleets built generation on site

Some operators built on-site generation rather than wait for a utility connection. This response shows that connection access was rationing deployment, not merely adding an ordinary procurement lead time. Utilities and regulators controlled the relevant process.

• Control, no measured capture
• Value moved

RESHUFFLE 06 - THE POWER SYSTEM ·Depot fleets ·2020-2026

Depot evidence does not describe home charging

This case concerns delivery fleets and heavy trucks, not households charging at home. Widely cited generation-interconnection queues concern power stations, not depot connections. The sources also tell us little about how electric fleets are financed, insured or resold.

Scope: depot energisation, transmission-scale load connections and power-generation connections are different processes with different queues.

ONE TECHNOLOGY. THREE CONSTRAINTS. ·US passenger ·2024-2026

United States: financing became the bottleneck

By 2025, US electric cars could handle the journeys buyers needed. The bottleneck had moved to the deal: lease payments, resale values, vehicle prices and tax support. Battery capability was no longer the main constraint in this market.

Scope: this explains the broad US passenger market, not every buyer or vehicle segment.

ONE TECHNOLOGY. THREE CONSTRAINTS. ·China passenger ·2020/21-2026

China: eligibility rules shaped competition

China had ample cell production, but supplier-qualification and domestic-content rules still decided which vehicles could compete. The evidence does not show whether charging or grid capacity also constrained Chinese sales.

ONE TECHNOLOGY. THREE CONSTRAINTS. ·Depot fleets ·2020-2026

Depot fleets: the connection controlled deployment

Fleet operators could buy trucks capable of completing their routes. Getting enough power to the depot could take nine months to three years, or five years when local upgrades were required. Utilities and regulators controlled that process.

ONE TECHNOLOGY. THREE CONSTRAINTS.

The constraint changed with the market and the job

US passenger sales depended on transaction economics. Chinese market access depended on supplier eligibility. Depot deployment depended on utility energisation. Value and control also moved separately: CATL earned strong profit while carmakers retained responsibility for the finished vehicle. Despite these different constraints, leading producers moved toward greater vertical integration. The evidence does not show whether their individual strategies caused that convergence or whether a common pressure did.

Scope: CATL's profit is audited. No supplier in this account replaced the carmaker as vehicle integrator.