The United Cancer Council is a charity that seeks, through affiliated local cancer societies, to encourage preventive and ameliorative approaches to cancer, as distinct from searching for a cure, which has been the emphasis of the older and better-known American Cancer Society, of which UCC is a splinter. The Internal Revenue Service revoked UCC's charitable exemption and the Tax Court upheld the revocation, precipitating this appeal.

So far as relates to this case, a charity, in order to be entitled to the charitable exemption from federal income tax, and to be eligible to receive tax-exempt donations, must be "organized and operated exclusively for ... [charitable] purposes" and "no part of the net earnings of [the charity may] inure[ ] to the benefit of any private shareholder or individual." 26 U.S.C. §§ 501(c)(3) (exemption); 170(c)(2)(B), (C) (receipt of donations); 26 C.F.R. § 1.501(a)-1(c), 1.501(c)(3)-1(d)(1)(i), (ii). The IRS claims that UCC (which is defunct) was not operated exclusively for charitable purposes, but rather was operated for, or also for, the private benefit of the fundraising company that UCC had hired, Watson & Hughey Company (W & H). The Service also claims that part of the charity's net earnings had inured to the benefit of a private shareholder or individual--W & H again. The Tax Court upheld the Service's second ground for revoking UCC's exemption--inurement--and did not reach the first ground, private benefit. The only issue before us is whether the court clearly erred, Fund for the Study of Economic Growth & Tax Reform v. IRS, 161 F.3d 755, 758-59 (D.C.Cir.1998); Church of Scientology v. Commissioner, 823 F.2d 1310, 1317 (9th Cir.1987); see also Walgreen Co. v. Commissioner, 68 F.3d 1006, 1009-10 (7th Cir.1995); Williams v. Commissioner, 1 F.3d 502, 505 (7th Cir.1993), in finding that a part of UCC's net earnings inured to the benefit of a private shareholder or individual.

It is important to understand what the IRS does not contend. It does not contend that any part of UCC's earnings found its way into the pockets of any members of the charity's board; the board members, who were medical professionals, lawyers, judges, and bankers, served without compensation. It does not contend that any members of the board were owners, managers, or employees of W & H, or relatives or even friends of any of W & H's owners, managers, or employees. It does not contend that the fundraiser was involved either directly or indirectly in the creation of UCC, or selected UCC's charitable goals. It concedes that the contract between charity and fundraiser was negotiated at an arm's length basis. But it contends that the contract was so advantageous to W & H and so disadvantageous to UCC that the charity must be deemed to have surrendered the control of its operations and earnings to the noncharitable enterprise that it had hired to raise money for it.

The facts are undisputed. In 1984, UCC was a tiny organization. It had an annual operating budget of only $35,000, and it was on the brink of bankruptcy because several of its larger member societies had defected to its rival, the American Cancer Society. A committee of the board picked W & H, a specialist in raising funds for charities, as the best prospect for raising the funds essential for UCC's survival. Another committee of the board was created to negotiate the contract. Because of UCC's perilous financial condition, the committee wanted W & H to "front" all the expenses of the fundraising campaign, though it would be reimbursed by UCC as soon as the campaign generated sufficient donations to cover those expenses. W & H agreed. But it demanded in return that it be made UCC's exclusive fundraiser during the five-year term of the contract, that it be given co-ownership of the list of prospective donors generated by its fundraising efforts, and that UCC be forbidden, both during the term of the contract and after it expired, to sell or lease the list, although it would be free to use it to solicit repeat donations. There was no restriction on W & H's use of the list. UCC agreed to these terms and the contract went into effect.

Over the five-year term of the contract, W & H mailed 80 million letters soliciting contributions to UCC. Each letter contained advice about preventing cancer, as well as a pitch for donations; 70 percent of the letters also offered the recipient a chance to win a sweepstake. The text of all the letters was reviewed and approved by UCC. As a result of these mailings, UCC raised an enormous amount of money (by its standards)--$28.8 million. But its expenses--that is, the costs borne by W & H for postage, printing, and mailing the letters soliciting donations, costs reimbursed by UCC according to the terms of the contract--were also enormous--$26.5 million. The balance, $2.3 million, the net proceeds of the direct-mail campaign, was spent by UCC for services to cancer patients and on research for the prevention and treatment of cancer. The charity was permitted by the relevant accounting conventions to classify $12.2 million of its fundraising expenses as educational expenditures because of the cancer information contained in the fundraising letters.

Although UCC considered its experience with W & H successful, it did not renew the contract when it expired by its terms in 1989. Instead, it hired another fundraising organization--with disastrous results. The following year, UCC declared bankruptcy, and within months the IRS revoked its tax exemption retroactively to the date on which UCC had signed the contract with W & H. The effect was to make the IRS a major creditor of UCC in the bankruptcy proceeding. The retroactive revocation did not, however, affect the charitable deduction that donors to UCC since 1984 had taken on their income tax returns. See Bob Jones University v. Simon, 416 U.S. 725, 728-29, 94 S. Ct. 2038, 40 L. Ed. 2d 496 (1974); Rev. Proc. 82-39, § 3.01, 1982-1 Cum. Bull. 759, 1982 WL 196338 (July 6, 1982).

The term "any private shareholder or individual" in the inurement clause of section 501(c)(3) of the Internal Revenue Code has been interpreted to mean an insider of the charity. Orange County Agricultural Society, Inc. v. Commissioner, 893 F.2d 529, 534 (2d Cir.1990); Church of Scientology v. Commissioner, supra, 823 F.2d at 1316-19; Church by Mail, Inc. v. Commissioner, 765 F.2d 1387, 1392 (9th Cir.1985); American Campaign Academy v. Commissioner, 92 T.C. 1053, 1066, 1989 WL 49678 (1989). A charity is not to siphon its earnings to its founder, or the members of its board, or their families, or anyone else fairly to be described as an insider, that is, as the equivalent of an owner or manager. The test is functional. It looks to the reality of control rather than to the insider's place in a formal table of organization. The insider could be a "mere" employee--or even a nominal outsider, such as a physician with hospital privileges in a charitable hospital, Harding Hospital, Inc. v. United States, 505 F.2d 1068, 1078 (6th Cir.1974); John E. Burke, "Hospital-Physician Joint Ventures," GCM 39862, 1991 WL 776308 (IRS Dec. 2, 1991), a licensor, Est of Hawaii v. Commissioner, 71 T.C. 1067, 1078-81, 1979 WL 3604 (1979), aff'd, 647 F.2d 170 (9th Cir.1981) (realistically, a case involving a founder's siphoning of charitable donations), or for that matter a fundraiser, National Foundation, Inc. v. United States, 13 Cl.Ct. 486, 494-95 (1987)--though the court in that case rejected the argument that the fundraiser controlled the charity.

The Tax Court's classification of W & H as an insider of UCC was based on the fundraising contract. Such contracts are common. Fundraising has become a specialized professional activity and many charities hire specialists in it. If the charity's contract with the fundraiser makes the latter an insider, triggering the inurement clause of section 501(c)(3) and so destroying the charity's tax exemption, the charity sector of the economy is in trouble. The IRS does not take the position that every such contract has this effect. What troubles it are the particular terms and circumstances of UCC's contract. It argues that since at the inception of the contract the charity had no money to speak of, and since, therefore, at least at the beginning, all the expenses of the fundraising campaign were borne by W & H, the latter was like a founder, or rather refounder (UCC was created in 1963), of the charity. The IRS points out that 90 percent of the contributions received by UCC during the term of the contract were paid to W & H to defray the cost of the fundraising campaign that brought in those contributions, and so argues that W & H was the real recipient of the contributions. It argues that because W & H was UCC's only fundraiser, the charity was totally at W & H's mercy during the five-year term of the contract--giving W & H effective control over the charity. UCC even surrendered the right to rent out the list of names of donors that the fundraising campaign generated. The terms of the contract were more favorable to the fundraiser than the terms of the average fundraising contract are.

Singly and together, these points bear no relation that we can see to the inurement provision. The provision is designed to prevent the siphoning of charitable receipts to insiders of the charity, not to empower the IRS to monitor the terms of arm's length contracts made by charitable organizations with the firms that supply them with essential inputs, whether premises, paper, computers, legal advice, or fundraising services.

Take the Service's first point, that W & H defrayed such a large fraction of the charity's total expenses in the early stages of the contract that it was the equivalent of a founder. Pushed to its logical extreme, this argument would deny the charitable tax exemption to any new or small charity that wanted to grow by soliciting donations, since it would have to get the cash to pay for the solicitations from an outside source, logically a fundraising organization. We can't see what this has to do with inurement. The argument is connected to another of the Service's points, that W & H was UCC's only fundraiser during the period of the contract. If UCC had hired ten fundraisers, the Service couldn't argue that any of them was so large a recipient of the charity's expenditures that it must be deemed to have controlled the charity. Yet in terms of the purposes of the inurement clause, it makes no difference how many fundraisers a charity employs. W & H obtained an exclusive contract, and thus was the sole fundraiser, not because it sought to control UCC and suck it dry, but because it was taking a risk; the exclusive contract lent assurance that if the venture succeeded, UCC wouldn't hire other fundraisers to reap where W & H had sown.

And it was only at the beginning of the contract period that W & H was funding UCC. As donations poured into the charity's coffers as a result of the success of the fundraising campaign, the charity began paying for the subsequent stages of the campaign out of its own revenues. True, to guarantee recoupment, the contract with W & H required UCC to place these funds in an escrow account, from which they could be withdrawn for UCC's charitable purposes only after W & H recovered the expenses of the fundraising campaign. But this is a detail; the important point is that UCC did not receive repeated infusions of capital from W & H. All the advances that W & H had made to UCC to fund the fundraising campaign were repaid. Indeed, it is an essential part of the government's case that W & H profited from the contract.

The other point that the Service makes about the exclusivity provision in the contract--that it put the charity at the mercy of the fundraiser, since if W & H stopped its fundraising efforts UCC would be barred from hiring another fundraiser until the contract with W & H expired--merely demonstrates the Service's ignorance of contract law. When a firm is granted an exclusive contract, the law reads into it an obligation that the firm use its best efforts to promote the contract's objectives. Wood v. Duff-Gordon, 222 N.Y. 88, 118 N.E. 214 (N.Y.1917) (Cardozo, J.); Advent Systems Ltd. v. Unisys Corp., 925 F.2d 670, 679-80 (3d Cir.1991); E. Allan Farnsworth, Contracts § 7.17, pp. 509-11 (3d ed.1999). If W & H folded its tent and walked away, it would be in breach of this implied term of the contract and UCC would be free to terminate the contract without liability.

The Service also misses the significance of the contract's asymmetrical treatment of the parties' rights in the donor list. The charitable-fundraising community distinguishes between "prospect files" and "housefiles." A prospect file is a list of people who have not given to the charity in question but are thought sufficiently likely to do so to be placed on the list of addressees of a direct-mail fundraising campaign. If the prospect responds with a donation, his or her name is transferred to the housefile, that is, the list of people who have made a donation to the charity. A housefile is very valuable, because people who have already donated to a particular charity are more likely to donate to it again than mere prospects are likely to donate to it for the first time. The housefile's value to the charity is thus as a list of people who are good prospects to respond favorably to future solicitations. Its value to the fundraiser is quite different. The fundraiser is not a charity. The value to it of a housefile that it has created is the possibility of marketing it (as a prospect file--but as a prospect file in which all the prospects are charitable donors rather than a mere cross-section of potential donors) to another charity that hires it. So it made perfect sense for the contract to give the fundraiser the exclusive right to use the UCC housefile that it created in raising money for other charities, while reserving to UCC the right to use the housefile to solicit repeat donations to itself.

The Service's point that has the most intuitive appeal is the high ratio of fundraising expenses, all of which went to W & H because it was UCC's only fundraiser during the term of the contract, to net charitable proceeds. Of the $28-odd million that came in, $26-plus million went right back out, to W & H. These figures are deceptive, because UCC got a charitable "bang" from the mailings themselves, which contained educational materials (somewhat meager, to be sure) in direct support of the charity's central charitable goal. A charity whose entire goal was to publish educational materials would spend all or most of its revenues on publishing, but this would be in support rather than in derogation of its charitable purposes.

Even if this point is ignored, the ratio of expenses to net charitable receipts is unrelated to the issue of inurement. For one thing, it is a ratio of apples to oranges: the gross expenses of the fundraiser to the net receipts of the charity. For all that appears, while UCC derived a net benefit from the contract equal to the difference between donations and expenses plus the educational value of the mailings, W & H derived only a modest profit; for we know what UCC paid it, but not what its expenses were. The record does contain a table showing that W & H incurred postage and printing expenses of $12.5 million, but there is nothing on its total expenses.

To the extent that the ratio of net charitable proceeds to the cost to the charity of generating those proceeds has any relevance, it is to a different issue, one not presented by this appeal, which is whether charities should be denied a tax exemption if their operating expenses are a very high percentage of the total charitable donations that they receive. To see that it's a different issue, just imagine that UCC had spent $26 million to raise $28 million but that the $26 million had been scattered among a host of suppliers rather than concentrated on one. There would be no issue of inurement, because the Service would have no basis for singling out one of these suppliers as being in "control" of UCC (or the suppliers as a group, unless they were acting in concert). But there might still be a concern either that the charity was mismanaged or that charitable enterprises that generate so little net contribution to their charitable goals do not deserve the encouragement that a tax exemption provides. Recall that most of UCC's fundraising appeals offered the recipient of the appeal a chance to win a sweepstake, a form of charitable appeal that, we are told, is frowned upon. There may even be a question of how reputable W & H is (or was). See Commonwealth v. Watson & Hughey Co., 128 Pa.Cmwlth. 484, 563 A.2d 1276 (Pa.Commw.1989). But these points go to UCC's sound judgment, not to whether W & H succeeded in wresting control over UCC from the charity's board.

UCC's low net yield is no doubt related to the terms of the fundraising contract, which were more favorable to the fundraiser than the average such contract. But so far as appears, they were favorable to W & H not because UCC's board was disloyal and mysteriously wanted to shower charity on a fundraiser with which it had no affiliation or overlapping membership or common ownership or control, but because UCC was desperate. The charity drove (so far as the record shows) the best bargain that it could, but it was not a good bargain. Maybe desperate charities should be encouraged to fold rather than to embark on expensive campaigns to raise funds. But that too is a separate issue from inurement. W & H did not, by reason of being able to drive a hard bargain, become an insider of UCC. If W & H was calling the shots, why did UCC refuse to renew the contract when it expired, and instead switch to another fundraiser?

We can find nothing in the facts to support the IRS's theory and the Tax Court's finding that W & H seized control of UCC and by doing so became an insider, triggering the inurement provision and destroying the exemption. There is nothing that corporate or agency law would recognize as control. A creditor of UCC could not seek the satisfaction of his claim from W & H on the ground that the charity was merely a cat's paw or alter ego of W & H, as in Pepper v. Litton, 308 U.S. 295, 311-12, 60 S. Ct. 238, 84 L. Ed. 281 (1939), or Freeman v. Complex Computing Co., 119 F.3d 1044, 1051-53 (2d Cir.1997). The Service and the Tax Court are using "control" in a special sense not used elsewhere, so far as we can determine, in the law, including federal tax law. It is a sense which, as the amicus curiae briefs filed in support of UCC point out, threatens to unsettle the charitable sector by empowering the IRS to yank a charity's tax exemption simply because the Service thinks the charity's contract with its major fundraiser too one-sided in favor of the fundraiser, even though the charity has not been found to have violated any duty of faithful and careful management that the law of nonprofit corporations may have laid upon it. The resulting uncertainty about the charity's ability to retain its tax exemption--and receive tax-exempt donations--would be a particular deterrent to anyone contemplating a donation, loan, or other financial contribution to a new or small charity. That is the type most likely to be found by the IRS to have surrendered control over its destiny to a fundraiser or other supplier, because it is the type of charity that is most likely to have to pay a high price for fundraising services. "Developments in the Law--Nonprofit Corporations," 105 Harv. L.Rev. 1578, 1649-51 (1992). It is hard enough for new, small, weak, or marginal charities to survive, because they are likely to have a high expense ratio, and many potential donors will be put off by that. The Tax Court's decision if sustained would make the survival of such charities even more dubious, by enveloping them in doubt about their tax exemption.

We were not reassured when the government's lawyer, in response to a question from the bench as to what standard he was advocating to guide decision in this area, said that it was the "facts and circumstances" of each case. That is no standard at all, and makes the tax status of charitable organizations and their donors a matter of the whim of the IRS.

There was no diversion of charitable revenues to an insider here, nothing that smacks of self-dealing, disloyalty, breach of fiduciary obligation or other misconduct of the type aimed at by a provision of law that forbids a charity to divert its earnings to members of the board or other insiders. What there may have been was imprudence on the part of UCC's board of directors in hiring W & H and negotiating the contract that it did. Maybe the only prudent course in the circumstances that confronted UCC in 1984 was to dissolve. Charitable organizations are plagued by incentive problems. Nobody owns the right to the profits and therefore no one has the spur to efficient performance that the lure of profits creates. Donors are like corporate shareholders in the sense of being the principal source of the charity's funds, but they do not have a profit incentive to monitor the care with which the charity's funds are used. Maybe the lack of a profit motive made UCC's board too lax. Maybe the board did not negotiate as favorable a contract with W & H as the board of a profitmaking firm would have done. And maybe tax law has a role to play in assuring the prudent management of charities. Remember the IRS's alternative basis for yanking UCC's exemption? It is that as a result of the contract's terms, UCC was not really operated exclusively for charitable purposes, but rather for the private benefit of W & H as well. Suppose that UCC was so irresponsibly managed that it paid W & H twice as much for fundraising services as W & H would have been happy to accept for those services, so that of UCC's $26 million in fundraising expense $13 million was the equivalent of a gift to the fundraiser. Then it could be argued that UCC was in fact being operated to a significant degree for the private benefit of W & H, though not because it was the latter's creature. That then would be a route for using tax law to deal with the problem of improvident or extravagant expenditures by a charitable organization that do not, however, inure to the benefit of insiders.

That in fact is the IRS's alternative ground for revoking the exemption, the one the Tax Court gave a bye to. It would have been better had the court resolved that ground as well as the inurement ground, so that the case could be definitively resolved in one appeal. But it did not, and so the case must be remanded to enable the court to consider it. We shall not prejudge the proceedings on remand. The usual "private benefit" case is one in which the charity has dual public and private goals, see, e.g., Better Business Bureau v. United States, 326 U.S. 279, 283, 66 S. Ct. 112, 90 L. Ed. 67 (1945); Living Faith, Inc. v. Commissioner, 950 F.2d 365 (7th Cir.1991) ; American Campaign Academy v. Commissioner, supra, 92 T.C. 1064-65, and that is not involved here. However, the board of a charity has a duty of care, just like the board of an ordinary business corporation, see, e.g., Riss v. Angel, 131 Wash.2d 612, 934 P.2d 669, 680-81 and n. 5 (Wash.1997); Fairhope Single Tax Corp. v. Rezner, 527 So. 2d 1232, 1236 (Ala.1987); Frances T. v. Village Green Owners Ass'n, 42 Cal. 3d 490, 229 Cal. Rptr. 456, 723 P.2d 573, 582 n. 13 (Cal.1986), and a violation of that duty which involved the dissipation of the charity's assets might (we need not decide whether it would--we leave that issue to the Tax Court in the first instance) support a finding that the charity was conferring a private benefit, even if the contracting party did not control, or exercise undue influence over, the charity. This, for all we know, may be such a case.

REVERSED AND REMANDED.