The traditional American car-buying system has a built-in middleman: the franchised dealership. Automakers manufacture vehicles, but in most states they cannot simply sell those vehicles directly to consumers.
Instead, cars move through independently owned dealers that maintain inventories, operate showrooms, employ sales and finance staff, advertise vehicles, and handle final transactions. That system carries a measurable cost.
A 2000 Goldman Sachs analysis estimated that shifting from the traditional franchised-dealer system to a direct-sales, build-to-order model could save about $2,225 per vehicle, based on an average vehicle price of $26,000 at the time.
That represented roughly 8.6% of the vehicle’s price. The estimate was later cited by the U.S. Department of Justice in an analysis of state laws restricting direct manufacturer sales.
However, the $2,225 figure should not be described simply as a dealer’s profit or markup. It represented potential distribution savings from eliminating or reducing costs associated with the traditional dealership intermediary, including inventory, facilities, sales personnel, and related expenses.
More than two decades later, the issue has become increasingly relevant as manufacturers experiment with direct-to-consumer sales.
A 2026 analysis from the International Center for Law & Economics (ICLE), using the Goldman Sachs framework and modern vehicle prices, estimates potential distribution savings of roughly $3,934 to $4,992 per vehicle. Those figures are an updated ICLE analysis, not a new Goldman Sachs study.
Where the $2,225 Estimate Came From
The Goldman Sachs analysis did not assume that the dealership’s entire gross profit could simply disappear under a direct-sales model.
Instead, it examined costs created by having vehicles pass through a traditional retail network. One major expense is inventory.
Under the conventional system, manufacturers produce vehicles before knowing exactly which customer will purchase them. Vehicles are shipped to dealers, where they may remain for weeks or months. Dealers must finance, insure, and store that inventory while paying for large lots and facilities.
A build-to-order model changes that equation. Instead of producing large quantities of vehicles and distributing them across thousands of dealerships, manufacturers can take customer orders and coordinate production and delivery more closely with actual demand.
The original Goldman Sachs framework estimated savings from reducing inventory-related costs along with other expenses associated with the dealer network.
The DOJ later cited the analysis, saying Goldman Sachs estimated approximately $2,225 in savings per vehicle, while identifying another roughly $1,000 in potential savings from product development, manufacturing flexibility, procurement, and supply improvements.
That does not mean every buyer would automatically receive a $2,225 discount. Some savings could instead become additional manufacturer profit or be spent on new direct-sales infrastructure.
Dealerships also require significant physical infrastructure. A conventional store needs a showroom, service department, vehicle storage, sales staff, managers, finance personnel, and administrative employees. It also requires enough land to hold substantial inventory.
A direct-sales system could reduce some of those requirements by moving much of the transaction online and using smaller delivery facilities.
Sales personnel represent another potential saving. Traditional dealers often employ salespeople, sales managers, and finance-and-insurance personnel. A fixed-price online purchasing process could reduce some of those expenses.
The argument is therefore not that dealerships perform no useful function. It is that manufacturers could potentially organize the same functions differently.
Why the Estimate Could Be Worth More Today
The original Goldman Sachs estimate was based on market conditions in 2000. Vehicle prices and dealership operating costs have changed dramatically since then.
ICLE’s 2026 analysis uses approximately $50,000 as a modern vehicle-price baseline and applies the original Goldman Sachs cost ratios. That produces estimated distribution savings of approximately $3,934 to $4,992 per vehicle, or about 7.9% to 9.9% of the assumed transaction price.
The modern calculation produces a larger dollar figure partly because vehicles themselves are considerably more expensive.
Inventory financing is another consideration. Dealers typically borrow money to maintain vehicle inventories, so carrying costs are affected by financing rates and how long vehicles remain unsold.
ICLE estimates roughly $1,105 in inventory-related savings per vehicle when applying the original Goldman Sachs ratio to a $50,000 vehicle. Sales-related expenses also contribute.
ICLE’s updated analysis estimates approximately $735 per vehicle in sales personnel savings using the Goldman Sachs framework.
Logistics represents another category, although the original Goldman Sachs estimate assigned it a relatively small share of the total savings.
A direct-sales system could potentially shorten the path between manufacturer and customer, especially when a vehicle is produced specifically for an identified buyer.
The growth of online shopping makes such a model more practical than it was in 2000. Customers can configure vehicles, arrange financing, and complete much of the purchase process without visiting a showroom. But the dealer network also provides services that manufacturers would still have to replace.
Customers need locations for test drives, vehicle delivery, repairs, and maintenance. A manufacturer operating directly would still have transportation, service, and customer support expenses.
That is why the Goldman Sachs figure should be viewed as an estimated opportunity for cost reduction rather than guaranteed consumer savings. There is also an important limitation to the modern calculation.
ICLE’s newer figures are not a fresh Goldman Sachs study. They apply the earlier analysis’s assumptions to current market conditions. The original research is more than 25 years old, so today’s actual savings could differ depending on inventory levels, labor costs, financing expenses, and how manufacturers structure their direct-sales operations.
Would Buyers Actually Save the Money?
That is the most important question. Removing the dealer intermediary does not automatically require manufacturers to reduce vehicle prices by the full amount of the estimated savings.
A manufacturer could retain some savings as profit, invest in online sales infrastructure, or use the money to reduce other operating expenses. There is also a counterargument from the dealership industry.
The National Automobile Dealers Association has challenged claims that eliminating franchised dealers would automatically generate all of the projected savings. Its research has argued that dealers provide services and investments that a direct-sales system would still need to reproduce.
That criticism matters because economic estimates depend on assumptions about which dealership functions can actually be eliminated.
Still, direct sales have demonstrated that a manufacturer can operate without the traditional franchise structure. Tesla has built a major direct-sales model, while other manufacturers have experimented with online purchasing and agency-style retail arrangements.
The economics, therefore, remain relevant. The original $2,225 estimate should not be interpreted as a $2,225 profit added to every vehicle by dealers. It represents potential distribution savings associated with reducing the costs of the traditional intermediary.
The updated ICLE analysis suggests that, under its assumptions, those potential savings could now reach approximately $4,000 to $5,000 per vehicle. Whether consumers would receive all of that amount is a separate matter.
What the research does show is that the traditional dealership system has real economic costs. The Goldman Sachs analysis estimated $2,225 per vehicle in potential savings from moving toward direct sales, and the DOJ subsequently used that work when examining the economic effects of state restrictions on manufacturer-to-consumer sales.
For today’s buyers, the debate is no longer simply about whether cars can be sold directly. It is about how much the traditional distribution structure costs and how much of that cost could realistically disappear under a different model.
The $2,225 figure is therefore best understood as a measure of potential efficiency, not a guaranteed discount, but it remains a useful benchmark for understanding why the role of the dealership intermediary continues to attract scrutiny.
The pricing structure also matters because removing the intermediary can change how a vehicle is sold. In a traditional dealership transaction, the final cost may involve negotiations over the vehicle price, dealer-installed accessories, financing options, and trade-in value.
A direct-sales model can replace much of that process with a fixed-price structure, making the transaction easier to understand for buyers.
However, direct sales do not eliminate every retail expense. Manufacturers still need customer-service operations, delivery networks, service facilities, and financing arrangements. The potential savings therefore depend on how efficiently those replacement systems are operated.
That is why the Goldman Sachs estimate remains a useful benchmark rather than a guaranteed discount.
