The district court granted summary judgment for the defendant, an auto dealer sued for violating the Truth in Lending Act by failing to disclose a finance charge that it levies on "subprime" purchasers of its used cars. A subprime purchaser is a purchaser whose credit rating is so poor that he, or in this case she, can obtain credit only from or through finance companies that specialize in high-risk borrowers. One of these finance companies, the one that financed the plaintiff's purchase of a used car from the defendant, charges the defendant a flat $400 per car financed, called a "holdback." The plaintiff contends that the defendant passes this charge on to its subprime customers but does not treat it as a finance charge in computing the annual percentage rate of interest that it discloses to them. In a standard holdback, the money held back is returned to the dealer if and when the purchaser pays off the loan. The parties do not discuss this wrinkle and for simplicity we'll assume the finance company never returns the money; nothing turns on the point. The dealer uses other finance companies that charge a different holdback, but, again for the sake of simplicity, we'll ignore that complication as well--which is in any event irrelevant to this plaintiff--and pretend that the holdback is always $400; actually, the average holdback paid by this dealer exceeds $700.

Were it true that the dealer tacked $400 (or some fraction thereof) onto the price of the cars it sold subprime purchasers, and did not tack the same amount onto the prices charged its other purchasers, the addition would indeed be a finance charge and the dealer would have to include it in computing the annual percentage rate of interest charged this class of credit customers. Walker v. Wallace Auto Sales, Inc., 155 F.3d 927 (7th Cir. 1998). But as long as the dealer raises the price to all its purchasers by the same amount to absorb this cost, so that the customer cannot avoid it by paying cash, it is not a finance charge within the meaning of the Truth in Lending Act. The Act's purpose is to enable consumers to decide whether or from whom to obtain credit, and a charge that does not affect the cost of one form of credit relative to another or to cash is irrelevant to that purpose. Balderos v. City Chevrolet, 214 F.3d 849, 851-52 (7th Cir. May 26, 2000).

The present case is intermediate between the two hypothetical variants that we have given. The defendant does not add a $400 charge to the price of cars sold its subprime customers, but there is evidence that it charges them a higher price on average than it charges its other customers, and the difference, the plaintiff argues, is a hidden finance charge. There is no reported appellate case quite like this. Balderos and Walker, on which the plaintiff relies, came to us in a critically different procedural posture. The complaint in each case alleged as in this one that the defendant added a finance charge to its credit sales and not to its cash sales, but in each case the district court dismissed the suit for failure to state a claim and so we were required to assume that the allegation was true. Here, by moving for summary judgment, the defendant forced the plaintiff to present evidence that this dealer really did include a secret finance charge in the sales price to its credit customers (actually a subset of those customers) but not in the price to its other customers. The plaintiff couldn't just stand on her complaint.

The dealer prices its used cars as follows. (It also sells new cars, but the plaintiff makes no argument regarding their pricing.) It takes the cash value of the car, either the trade-in value or the price the car would command at an auction, and adds the cost of any repairs the dealer has made. To the sum of these two items--we'll call that sum the dealer's "cost of car"--the dealer adds a uniform markup of $5,700. The sum of the cost of car and the markup is the dealer's list price for the used car. The dealer's salesmen are not expected to sell most, perhaps any, cars at list price; that price is just the beginning of the negotiation. The salesman tries to get as much as he can, of course, but his commission is a percentage not of the sales price but of the defendant's net profit on the sale. The net profit is the actual sales price (which must be approved by the defendant's business office) minus not only what we're calling the "cost of car" but also $700, representing an allocation of the dealer's overhead, and, in the case of subprime purchasers, the "holdback"--in this case the $400 that the finance company charged the dealer for financing the plaintiff's purchase.

This method of computing the salesman's commissions implies that for him to get the identical commission on two otherwise identical cars, one sold to a subprime purchaser and the other to a prime or cash purchaser, the price to the subprime purchaser would have to be $400 higher than the price to the other purchaser. Otherwise the dealer's net profit would be less on the car sold to the subprime purchaser and so the salesman's commission would also be less. It does not follow, however, as the plaintiff seems to believe, that the price to the subprime purchaser would be $400 higher, or for that matter one penny higher. The salesman will try in every negotiation to strike the hardest bargain he can, whether or not there is a holdback. The plaintiff paid $8,800 for the car she bought from the defendant. Since she was willing to pay that much, it would have been irrational to charge her less if it were discovered that she wasn't a subprime purchaser after all. The only significance of such a discovery would be to reveal that the sale at $8,800 was more profitable than the dealer had thought and that the salesman was entitled to a higher commission.

This example shows how unlikely it is that the holdback is passed on to subprime purchasers. If there were no holdbacks at all, the defendant's cost of doing business would be lower, and conceivably this would induce it to reduce the $5,700 markup or to permit its salesmen to negotiate somewhat lower sale prices, since generally the lower a company's costs the lower its profit- maximizing price (even if it's a monopolist, which there is no reason to think this dealer is). But there is no reason to suppose that subprime purchasers would be particular beneficiaries. The defendant would still be trying to get as close to its list prices as it could with all its purchasers.

The only situation in which the $400 holdback would be demonstrably a hidden finance charge would be where the defendant would have sold the car to a cash purchaser for less than $400 over the defendant's cost of car. Suppose that cost were $2,000. It would make no sense for the defendant to sell to a subprime purchaser for less than $2,400, because in that event the defendant would net less than $2,000 and it could get more by auctioning the car or trading it in for another car, thus avoiding the $400 cost that the finance company charges it for financing a sale to a subprime purchaser. (We're ignoring repair costs in this example.) Yet the dealer might sell the same car to a hard-bargaining nonsubprime purchaser for between $2,000 and $2,400, if it didn't think it could get more than $2,400 from anyone else. For a sale at any price between these figures would cover the cost of making the sale, a cost that, in the case of a sale not to a subprime purchaser, does not include a $400 charge by the finance company. The $700 in overhead costs that the defendant allocates to every sale is incurred whether or not a car is sold, and so it is not saved, as the holdback is, by refusing to sell the car at a price that does not cover that cost. See Autotrol Corp. v. Continental Water Systems Corp., 918 F.2d 689, 692-93 (7th Cir. 1990). That is why it does not figure in our example of the minimum price that the dealer would charge to a subprime purchaser and to another purchaser, respectively.

This case is unlike the example, as is obvious from the price that the plaintiff paid. That price included a markup greatly in excess of $400--the dealer's net profit on the sale to her was $2,599.15. There is no evidence of a sale by the defendant to anyone at a markup below $400. The plaintiff did present evidence that, on average, subprime purchasers pay higher prices than the dealer's other purchasers, after correction for differences in the cost of car. But there is no evidence that the higher price reflects an effort by thedefendant to stick such purchasers with a $400 finance charge. We have seen that such a policy would be irrational except in the case of a very low (below $400) markup, of which there is no evidence.

Further undermining the plaintiff's theory of liability is the fact that the record reveals that prime credit purchasers from the defendant pay, on average, higher prices--before the cost of credit is computed--than cash purchasers, even though there is no holdback in the case of prime credit purchasers. Although the dealer's average markup (net of the overhead allocation) on holdback transactions is $1,800 and on cash transactions only $900, its average markup on credit transactions in which there is no holdback is $1,300. The implication is that cash customers are more credible or savvy bargainers than credit customers, and that credit customers with good credit ratings are more credible or savvy bargainers than credit customers with bad credit ratings (that is, the subprime purchasers). The record is consistent with this conjecture; the plaintiff accepted the price offered by the defendant's salesman, with no attempt to bargain him down, as she very well might have done since the markup to her (which, net of overhead, was approximately $3,000) was more than twice the dealer's average markup for nonsubprime credit customers and more than three times its average markup for cash customers. There is no evidence that she paid $400 more, or for that matter one cent more, than she would have paid had she not been a subprime purchaser--no evidence, in short, that a finance charge was "buried" in the cash price of the car. That kills her case. See Walker v. Wallace Auto Sales, Inc., supra, 155 F.3d at 933-34 and n. 9.

Courts in other contexts, notably antitrust enforcement, have been reluctant to get into issues of "passing on." E.g., Illinois Brick Co. v. Illinois, 431 U.S. 720 (1977). Such issues tend to be intractable to the methods of litigation. It is possible though unlikely that holdbacks influence the pricing of the defendant's cars to subprime purchasers, but to explore the possibility would enormously complicate the litigation of Truth in Lending Act claims, would cast a large cloud of potential liability over used-car dealers and other firms that use list prices as just the starting point for negotiation, and by doing so would make the class action a truly fearsome instrument of consumer-finance litigation. This case began as a class action; the consumer- finance class action is the specialty of the law firm representing the plaintiff.

This is not a case in which a charge that is really a finance charge is called something else. It is a case in which the plaintiff seeks to track a cost incurred by the dealer into the prices charged the dealer's customers. The plaintiff has, as we have seen, failed to do that. Her failure suggests the futility of a "passing on" theory of liability under the Truth in Lending Act.

The complaint charges that the defendant's failure to figure the holdback into the annual percentage rate of interest also violated the Illinois Consumer Fraud Act, 815 ILCS 505/2. That charge was rightly dismissed too, because compliance with the disclosure requirements in the federal Truth in Lending Act is a defense under the Illinois act. See Lanier v. Associates Finance, Inc., 499 N.E.2d 440, 447 (Ill. 1986).

Affirmed.