Norby Walters, who represents entertainers, tried to move into the sports business. He signed 58 college football players to contracts while they were still playing. Walters offered cars and money to those who would agree to use him as their representative in dealing with professional teams. Sports agents receive a percentage of the players’ income, so Walters would profit only to the extent he could negotiate contracts for his clients. The athletes’ pro prospects depended on successful completion of their collegiate careers. To the NCAA, however, a student who signs a contract with an agent is a professional, ineligible to play on collegiate teams. To avoid jeopardizing his clients’ careers, Walters dated the contracts after the end of their eligibility and locked them in a safe. He promised to lie to the universities in response to any inquiries. Walters inquired of sports lawyers at Shea & Gould whether this plan of operation would be lawful. The firm rendered an opinion that it would violate the NCAA’s rules but not any statute.
Having recruited players willing to fool their universities and the NCAA, Walters discovered that they were equally willing to play false with him. Only 2 of the 58 players fulfilled their end of the bargain; the other 56 kept the cars and money, then signed with other agents. They relied on the fact that the contracts were locked away and dated in the future, and that Walters’ business depended on continued-secrecy, so he could not very well sue to enforce their promises. When the 56 would neither accept him as their representative nor return the payments, Walters resorted to threats. One player, Maurice Douglass, was told that his legs would be broken before the pro draft unless he repaid Walters’ firm. A 75-page indictment charged Walters and his partner Lloyd Bloom with conspiracy, RICO violations (the predicate felony was extortion), and mail fraud. The fraud: causing the universities to pay scholarship funds to athletes who had become ineligible as a result of the agency contracts. The mail: each university required its athletes to verify their eligibility to play, then sent copies by mail to conferences such as the Big Ten.
. After a month-long trial and a week of deliberations, the jury convicted Walters and Bloom. We reversed, holding that the district judge had erred in declining to instruct the jury that reliance on Shea & Gould’s advice could prevent the formation of intent to defraud the universities. 913 F.2d 388, 391-92 (1990). Any dispute about the adequacy of Walters’ disclosure to his lawyers and the bona tides of his reliance was for the jury, we concluded. Because Bloom declined to waive his own attorney-client privilege, we held that the defendants must be retried separately. Id. at 392-93. On remand, Walters asked the district court to dismiss the indictment, arguing that the evidence presented at trial is insufficient to support the convictions. After the judge denied this motion, 775 F.Supp. 1173 (N.D.Ill.1991), Walters agreed to enter a conditional Alford plea: he would plead guilty to mail fraud, conceding that the record of the first trial supplies a factual basis for a conviction while reserving his right to contest the sufficiency of that evidence. In return, the prosecutor agreed to dismiss the RICO and conspiracy charges *1222and to return to Walters all property that had been forfeited as a result of his RICO conviction. Thus a case that began with a focus on extortion has become a straight mail fraud prosecution and may undergo yet another transformation. The prosecutor believes that Walters hampered the investigation preceding his indictment. See In re Feldberg, 862 F.2d 622 (7th Cir.1988) (describing some of the investigation). The plea agreement reserves the prosecutor’s right to charge Walters with perjury and obstruction of justice if we should reverse the conviction for mail fraud.
“Whoever, having devised ... any scheme or artifice to defraud, or for obtaining money or property by means of false or fraudulent pretenses, representations, or promises ... places in any post office or authorized depository for mail matter, any matter or thing whatever to be sent or delivered by the Postal Service ... or knowingly causes [such matter or thing] to be delivered by mail” commits the crime of mail fraud. 18 U.S.C. § 1341. Norby Walters did not mail anything or cause anyone else to do so (the universities were going to collect and mail the forms no matter what Walters did), but the Supreme Court has expanded the statute beyond its literal terms, holding that a mailing by a third party suffices if it is “incident to an essential part of the scheme,” Pereira v. United States, 347 U.S. 1, 8, 74 S.Ct. 358, 363, 98 L.Ed. 435 (1954). While stating that such mailings can turn ordinary fraud into mail fraud, the Court has cautioned that the statute “does not purport to reach all frauds, but only those limited instances in which the use of the mails is a part of the execution of the fraud”. Kann v. United States, 323 U.S. 88, 95, 65 S.Ct. 148, 151, 89 L.Ed. 88 (1944). Everything thus turns on matters of degree. Did the schemers foresee that the mails would be used? Did the mailing advance the success of the scheme? Which parts of a scheme are “essential”? Such questions lack obviously right answers, so it is no surprise that each side to this ease can cite several of our decisions in support. Compare United States v. McClain, 934 F.2d 822, 835 (7th Cir.1991), and United States v. Kwiat, 817 F.2d 440, 443-44 (7th Cir.1987), among cases reversing convictions because use of the mails was too remote or unforeseeable, with Messinger v. United States, 872 F.2d 217 (7th Cir.1989), among many cases holding that particular uses of the mails were vital to the scheme and foreseeable.
“The relevant question ... is whether the mailing is part of the execution of the scheme as conceived by the perpetrator at the time”. Schmuck v. United States, 489 U.S. 705, 715, 109 S.Ct. 1443, 1450, 103 L.Ed.2d 734 (1989). Did the evidence establish that Walters conceived a scheme in which mailings played a role? We think not — indeed, that no reasonable juror could give an affirmative answer to this question. Walters hatched a scheme to make money by taking a percentage of athletes’ pro contracts. To get clients he signed students while college eligibility remained, thus avoiding competition from ethical agents. To obtain big pro contracts for these clients he needed to keep the deals secret, so the athletes could finish their collegiate careers. Thus deceit was an ingredient of the plan. We may assume that Walters knew that the universities would ask athletes to verify that they were eligible to compete as amateurs. But what role do the mails play? The plan succeeds so long as the athletes conceal their contracts from their schools (and remain loyal to Walters). Forms verifying eligibility do not help the plan succeed instead they create a risk that it will be discovered if a student should tell the truth. Cf. United States v. Maze, 414 U.S. 395, 94 S.Ct. 645, 38 L.Ed.2d 603 (1974). And it is the forms, not their mailing to the Big Ten, that pose the risk. For all Walters cared, the forms could sit forever in cartons. Movement to someplace else was irrelevant. In Schmuck, where the fraud was selling cars with rolled-back odometers, the mailing was essential to obtain a new and apparently “clean” certificate of title; no certificates of title, no marketable cars, no hope for success. Even so, the Court divided five to four on the question whether the mailing was sufficiently integral to the scheme. A college’s mailing to its conference has less to do with the plot’s success than the mailings that transferred title in Schmuck.
*1223To this the United States responds that the mailings were essential because, if a college had neglected to send the athletes’ forms to the conference, the NCAA would have barred that college’s team from compete ing. Lack of competition would spoil the athletes’ pro prospects. Thus the use of the mails was integral to the profits Walters hoped to reap, even though Walters would have been delighted had the colleges neither asked any questions of the athletes nor put the answers in the mail. Let us take this as sufficient under Schmuck (although we have our doubts). The question remains whether Walters caused the universities to use the mails.1 A person “knowingly causes” the use of the mails when he “acts with the knowledge that the use of the mails will follow in the ordinary course of business, or where such use can reasonably be foreseen.” United States v. Kuzniar, 881 F.2d 466, 472 (7th Cir.1989), quoting Pereira, 347 U.S. at 8-9, 74 S.Ct. at 363. The paradigm is insurance fraud. Perkins tells his auto insurer that his car has been stolen, when in fact it has been sold. The local employee mails the claim to the home office, which mails a check to Perkins. Such mailings in the ordinary course of business are foreseeable. E.g., United States v. Richman, 944 F.2d 323 (7th Cir.1991). Similarly, a judge who takes a bribe derived from the litigant’s bail money causes the use of the mails when the ordinary course is to refund the bond by mail. E.g., United States v. Murphy, 768 F.2d 1518, 1529-30 (7th Cir.1985). The prosecutor contends that the same approach covers Walters.
No evidence demonstrates that Walters actually knew that the colleges would mail the athletes’ forms. The record is barely sufficient to establish that Walters knew of the forms’ existence; it is silent about Walters’ knowledge of the forms’ disposition. The only evidence implying that Walters knew that the colleges had students fill out forms is an ambiguous reference to “these forms” in the testimony of Robert Perryman. Nothing in the record suggests that Perryman, a student athlete, knew what his university did with the forms, let alone that Perryman passed this information to Walters. So the prosecutor is reduced to the argument that mailings could “reasonably be foreseen.” Yet why should this be so? Universities frequently collect information that is stashed in file drawers. Perhaps the NCAA just wants answers available for inspection in the event a question arises, or. the university wants the information for its own purposes (to show that it did not know about any improprieties that later come to light). What was it about these forms that should have led a reasonable person to foresee their mailing? Recall that Walters was trying to break into the sports business. Counsel specializing in sports law told him that his plan would not violate any statute. These lawyers were unaware of the forms (or, if they knew about the forms, were unaware that they would be mailed). The prosecutor contends that Walters neglected to tell his lawyers about the eligibility forms, spoiling their opinion; yet why would Walters have to brief an expert in sports law if mailings were foreseeable even to a novice?
In the end, the prosecutor insists that the large size and interstate nature of the NCAA demonstrate that something would be dropped into the mails. To put this only slightly differently, the prosecutor submits that all frauds involving big organizations necessarily are mail frauds, because big organizations habitually mail things. No evidence put before the jury supports such a claim, and it is hardly appropriate for judicial notice in a criminal case. Moreover, adopting this perspective would contradict the assurance of Kann, 323 U.S. at 95, 65 S.Ct. at 151, and many later cases that most frauds are covered by state law rather than § 1341. That statute has been expanded considerably *1224by judicial interpretation, but it does not make a federal crime of every deceit. The prosecutor must prove that the use of the mails was foreseeable, rather than calling on judicial intuition to repair a rickety case.
There is a deeper problem with the theory of this prosecution. The United States tells us that the universities lost their scholarship money. Money is property; this aspect of the prosecution does not encounter a problem under McNally v. United States, 483 U.S. 350, 107 S.Ct. 2875, 97 L.Ed.2d 292 (1987). Walters emphasizes that the universities put his 58 athletes on scholarship long before he met them and did not pay a penny more than they planned to do. But a jury could conclude that had Walters’ clients told the truth, the colleges would have stopped their scholarships, thus saving money. So we must assume that the universities lost property by reason of Walters’ deeds. Still, they were not out of pocket to Walters; he planned to profit by taking a percentage of the players’ professional incomes, not of their scholarships. Section 1341 condemns “any scheme or artifice to' defraud, or for obtaining money or property” (emphasis added). If the universities were the victims, how did he “obtain” their property?, Walters asks.
According to the United States, neither an actual nor a potential transfer of property from the victim to the defendant is essential. It is enough that the victim lose; what (if anything) the schemer hopes to gain plays no role in the definition of the offense. We asked the prosecutor at oral argument whether on this rationale practical jokes violate § 1341. A mails B an invitation to a surprise party for their mutual friend C. B drives his car to the place named in the invitation. But there is no party; the address is a vacant lot; B is the butt of a joke. The invitation came by post; the cost of gasoline means that B is out of pocket. The prosecutor said that this indeed violates § 1341, but that his office pledges to use prosecutorial discretion wisely. Many people will find this position unnerving (what if the prosecutor’s policy changes, or A is politically unpopular and the prosecutor is looking for a way to nail him?). Others, who obey the law out of a sense of civic obligation rather than the fear of sanctions, will alter their conduct no matter what policy the prosecutor follows. Either way, the idea that practical jokes are federal felonies would make a joke of the Supreme Court’s assurance that § 1341 does not cover the waterfront of deceit.
Practical jokes rarely come to the attention of federal prosecutors, but large organizations are more successful in gaining the attention of public officials. In this case the mail fraud statute has been invoked to shore up the rules of an influential private association. Consider a parallel: an association of manufacturers of plumbing fixtures adopts a rule providing that its members will not sell “seconds” (that is, blemished articles) to the public. The association proclaims that this rule protects consumers from shoddy goods. To remain in good standing, a member must report its sales monthly. These reports flow in by mail. One member begins to sell “seconds” but reports that it is not doing so. These sales take business away from other members of the association, who lose profits as a result. So we have mail, misrepresentation, and the loss- of property, but the liar does not-get any of the property the other firms lose. Has anyone committed a federal crime? The answer is yes — but the statute is the Sherman Act, 15 U.S.C. § 1, and the perpetrators are the firms that adopted the “no seconds” rule. United States v. Trenton Potteries Co., 273 U.S. 392, 47 S.Ct. 377, 71 L.Ed. 700 (1927). The trade association we have described is a cartel, which the firm selling “seconds” was undermining. Cheaters depress the price, causing the monopolist to lose money. Typically they go to great lengths to disguise their activities, the better to increase their own sales and avoid retaliation. The prosecutor’s position in our case would make criminals of the cheaters, would use § 1341 to shore up cartels.
Fanciful? Not at all. Many scholars understand the NCAA as a cartel, having power in the market for athletes. E.g., Arthur A Fleisher III, Brian L. Goff & Robert D. Tollison, The National Collegiate Athletic Association: A Study in Cartel Behavior (1992); Joseph P. Bauer, Antitrust and Sports: Must Competition on the Field Displace Competition in the Marketplace?, 60 *1225Tenn.L.Rev. 263 (1993); Roger D. Blair & Jeffrey L. Harrison, Cooperative Buying, Monopsony Power, and Antitrust Policy, 86 Nw.U.L.Rev. 331 (1992); Lee Goldman, Sports and Antitrust: Should College Students be Paid to Play?, 65 Notre Dame L.Rev. 206 (1990); Richard B. McKenzie & E. Thomas Sullivan, Does the NCAA Exploit College Athletes? An Economic and Legal Reinterpretation, 32 Antitrust Bull. 373 (1987); Stephen F. Ross, Monopoly Sports Leagues, 73 Minn.L.Rev. 643 (1989). See also NCAA v. University of Oklahoma, 468 U.S. 85, 104 S.Ct. 2948, 82 L.Ed.2d 70 (1984) (holding that the NCAA’s arrangements for the telecasting of college football violated the Sherman Act); Banks v. NCAA, 977 F.2d 1081 (7th Cir.1992) (showing disagreement among members of this court whether the NCAA’s restrictions on athletes violate the Sherman Act). The NCAA depresses athletes’ income — restricting payments to the value of tuition, room, and board, while receiving services. of substantially greater worth. The NCAA treats this as desirable preservation of amateur sports; a more jaundiced eye would see it as the use of monopsony power to obtain athletes’ services for less than the competitive market price. Walters then is cast in the role of a cheater, increasing the payments to the student athletes. Like other cheaters, Walters found it convenient to hide his activities. If, as the prosecutor believes, his repertory included extortion, he has used methods that the law denies to persons fighting cartels, but for the moment we are concerned only with the deceit that caused the universities to pay stipends to “professional” athletes. For current purposes it matters not whether the NCAA actually monopsonies the market for players; the point of this discussion is that the prosecutor’s theory makes criminals of those who consciously cheat on the rules of a private organization, even if that organization is a cartel. We pursue this point because any theory that makes criminals of cheaters raises a red flag.
Cheaters are not self-conscious champions of the public weal. They are in it for profit, as rapacious and mendacious as those who hope to collect monopoly rents. Maybe more; often members of cartels believe that monopoly serves the public interest, and they take their stand on the platform of business ethics, e.g., National Society of Professional Engineers v. United States, 435 U.S. 679, 98 S.Ct. 1355, 55 L.Ed.2d 637 (1978), while cheaters’ glasses have been washed with cynical acid. Only Adam Smith’s invisible hand turns their self-seeking activities to public benefit. It is cause for regret if prosecutors, assuming that persons with low regard for honesty must be villains, use the criminal laws to suppress the competitive process that undermines cartels. Of course federal laws have been used to enforce cartels before; the Federal Maritime Commission is a cartel-enforcement device. Inconsistent federal laws also occur; the United States both subsidizes tobacco growers and discourages people from smoking. So if the United States simultaneously forbids cartels and forbids undermining cartels by cheating, we shall shrug our shoulders and enforce both laws, condemning practical jokes along the way. But what is it about § 1341 that labels as a crime all deceit that inflicts any loss on anyone? Firms often try to fool their competitors, surprising them with new products that enrich their treasuries at their rivals’ expense. Is this mail fraud because large organizations inevitably use the mail? “[A]ny scheme or artifice to defraud, or for obtaining money or property by means of false or fraudulent pretenses, representations, or promises” reads like a description of schemes to get money or property by fraud rather than methods of doing business that incidentally cause losses.
None of the Supreme Court’s mail fraud cases deals with a scheme in which the defendant neither obtained nor tried to obtain the victim’s property. It has, however, addressed the question whether 18 U.S.C. § 371, which prohibits conspiracies to defraud the United States, criminalizes plans that cause incidental loss to the Treasury. Tanner v. United States, 483 U.S. 107, 130, 107 S.Ct. 2739, 2752, 97 L.Ed.2d 90 (1987), holds that § 371 applies only when the United States is a “target” of the fraud; schemes that cause indirect losses do not violate that statute. McNally tells us that § 371 covers a broader range of frauds than does § 1341, *1226see 483 U.S. at 358-59 n. 8, 107 S.Ct. at 2881, n. 8, and it follows that business plans causing incidental losses are not mail fraud. We have been unable to find any appellate cases squarely resolving the question whether the victim’s loss must be an objective of the scheme rather than a byproduct of it, perhaps because prosecutions of the kind this case represents are so rare.2 According to the prosecutor, however, there have been such cases, and in this circuit. The United States contends that we have already held that a scheme producing an incidental loss violates § 1341. A representative sample of the cases the prosecutor cites shows that we have held no such thing.
For example, United States v. Ashman, 979 F.2d 469, 477-83 (7th Cir.1992), affirms convictions for fraud in a matched-order scheme on the floor of the Chicago Board of Trade. Customers sent orders for execution “at the market.” Traders paired some of these orders off the market at times chosen to divert profits to themselves. This deprived the customers of the benefits provided by an open-outcry auction; more important, it moved money directly from customers’ accounts to the traders’ accounts. The transfers were the objective of the scheme. Nothing comparable took place here; no money moved from the universities to Walters. There is, however, some parallel in Ashman: we held that trades executed after the market was limit-up or limit-down could not support mail fraud convictions. Id. at 479. Once the market had moved the limit for the day, customers received the same price no matter when or with whom they traded. A customer willing to trade at a known price is like a university willing to give a scholarship to a known athlete. A customer who loses the honesty of the traders, but no money, has not been defrauded of property; a university that loses the benefits of amateurism likewise has been deprived only of an intangible right, which per McNally does not support a conviction.3
United States v. Richman sustained mail fraud convictions arising out of a lawyer’s attempt to bribe the claims adjuster for an insurance company. Retained to represent the victim of an accident, the lawyer offered the adjuster 5% of any settlement. Here was a fraud aimed at obtaining money from the insurer — a settlement was the objective rather than a byproduct of the scheme. The lawyer defended by contending that, because his client really had been injured, the insurer would have paid anyway. 944 F.2d at 330. This was highly implausible; why was the lawyer willing to bribe the adjuster? The scheme was designed to increase the settlement, or to induce the insurer to pay even if its insured was not negligent. At all events, we observed, the statute prohibits schemes that are designed to bilk other persons out of money or other property; the lawyer’s scheme had this objective even though the deceit might have been unnecessary. 944 F.2d at 330-31. Walters lacked any similar design to separate the universities from their money. Then there is United States v. Jones, 938 F.2d 737 (7th Cir.1991). The Joneses impersonated loan brokers. In exchange for $4,000 paid up front, they prom*1227ised to procure large loans for their clients. Although they accepted almost $10 million in fees, they never found funding for a single client. Next they dispersed the money in an effort to prevent the assessment and collection of taxes on the booty. Our holding that the scam violated both § 371 and § 1343 (wire fraud) by preventing the United States from taking its cut of the proceeds does not support a conclusion that any deceit that incidentally causes a loss to someone also violates federal law.
Many of our cases ask whether a particular scheme deprived a victim of property. E.g., Lombardo v. United States, 865 F.2d 155, 159-60 (7th Cir.1989). They do so not with an emphasis on “deprive” but with an emphasis on “property” — which, until the enactment of 18 U.S.C. § 1346 after Walters’ conduct, was essential to avoid the “intangible rights” doctrine that McNally jettisoned. No one doubted that the schemes were designed to enrich the perpetrators at the victims’ expense; the only difficulty was the proper characterization of the deprivation.4 Not until today have we dealt with a scheme in which the defendants’ profits were to come from legitimate transactions in the market, rather than at the expense of the victims. Both the “scheme or artifice to defraud” clause and the “obtaining money or property” clause of § 1343 contemplate a transfer of some kind. Accordingly, following both the language of § 1341 and the implication of Tanner, we hold that only a scheme to obtain money or other property from the victim by fraud violates § 1341. A deprivation is a necessary but not a sufficient condition of mail fraud. Losses that occur as byproducts of a deceitful scheme do not satisfy the statutory requirement.
Anticipating that we might come to this conclusion, the prosecutor contends that Walters is nonetheless guilty as an aider and abettor. If Walters did not defraud the universities, the argument goes, then the athletes did. Walters put them up to it and so is guilty under 18 U.S.C. § 2, the argument concludes. But the indictment charged a scheme by Walters to defraud; it did not depict Walters as an aide de camp in the students’ scheme. The jury received a boilerplate § 2 instruction; this theory was not argued to the jury, or for that matter to the district court either before or after the remand. Independent problems dog this recasting of the scheme — not least the difficulty of believing that the students hatched a plot to employ fraud to receive scholarships that the universities had awarded them long before Walters arrived on the scene, and the lack of evidence that the students knew about or could foresee any mailings. Walters is by all accounts a nasty and untrustworthy fellow, but the prosecutor did not prove that his efforts to circumvent the NCAA’s rules amounted to mail fraud.
REVERSED.