UNITED STATES of America, Plaintiff-Appellee, v. Georgia L. THOMPSON, Defendant-Appellant, 484 F.3d 877

EASTERBROOK, Chief Judge. In 2005 Wisconsin selected Adelman Travel'Group as its travel agent for about 40% of its annual travel budget of $75 million. The selection came after an elaborate process presided over by Georgia Thompson, a

section chief in the state’s Bureau of Procurement. Statutes and regulations require procurement decisions to be made on the basis of cost and service rather than politics. Wis. Stat. §§ 16.70-16.78; Wis. Admin. Code § 10.08. Thompson steered the contract to Adelman Travel, the low bidder, even though other members of the selection group rated its rivals more highly. A jury convicted Thompson of violating 18 U.S.C. § 666 and § 1341. The prosecution’s theory was that any politically motivated departure from state administrative rules is a federal crime, when either the mails or federal funds are involved.

Thompson was sentenced to 18 months’ imprisonment and compelled to begin serving that term while her appeal was pending. After concluding that Thompson is innocent, we reversed her conviction so that she could be released. This opinion is the explanation that our order of April 5 promised. Adelman Travel was the low bidder, but a low price for lousy service is no bargain. Wisconsin’s rules give price only a 25% weight (300 of 1200 points) in the selection process. About 58% (700 points) goes to service, which a working group evaluates subjectively based on written presentations.

Adelman had the second-best score for service; Omega World Travel came in third. The combined price-and-service rating had Adelman in the lead. (Fox World Travel received the best service score but had a noncompetitive price.) The final 17% of the score (200 points) depends on the working group’s assessment of oral presentations.

These presentations (often dubbed “beauty contests” or “dog-and-pony shows” that may reward the flashiest PowerPoint slides) need not be related to either price or the pitchman’s probable quality of service; why the state gives them any weight, independent of price or quality, is a mystery, but not one we need unravel. Adelman Travel must have made a bad presentation, for six of the seven members of the working group gave it poor marks (from a low of 120 points to a high of 165), while awarding Omega scores between 155 and 200. Thompson alone gave Adelman a higher score (185 for Adelman, 160 for Omega).

Adelman Travel’s disastrous oral presentation left Omega World Travel with the highest total score. The prosecution’s theory is that Omega should have received the contract on the spot but that for political reasons Thompson ordered a delay. Thompson told her colleagues that a decision for Omega, which is based on the East Coast, would not go over well with her boss, Pat Farley.

A jury also could conclude that Thompson said something to the effect that for “political reasons” Adelman Travel had to get this contract. (Witnesses related different versions of what Thompson said, but in each account “politics” or “political” played some role.) Thompson tried to engage in logrolling, offering to change her scores for bidders on other travel contracts if members of the working group would change their scores on this contract.

Horse-trading proved to be unacceptable to the selection group, but a member other than Thompson suggested that the contract be rebid on a best-and-final basis, as state law permitted. Wis. Stat. § 16.72(2m)(e), (g). Adelman Travel reduced its price, which — keeping all other elements of the score constant — left Adel-man and Omega with 1027 points apiece. The tie depended on rounding to the nearest whole number. Adelman Travel’s score was 1026.6, while Omega World Travel’s score was 1027.3.

After Thompson (with her supervisors’ consent) deemed the contest a draw — sensibly, as the difference was trivial compared to the amount of subjectivity and variance in the committee members’ evaluations — Thompson employed a tie-breaking procedure, specified by state law, that gave weight to items not previously figured into the price comparison and declared Adelman Travel to be the winner. The prosecutor contends that this episode played a role in the Bureau of Procurement’s decision three months later to give Thompson a $1,000 raise in her annual salary.

Post hoc ergo propter hoc is the name of a logical error, not a reason to infer causation. But Thompson does not contend that the evidence was insufficient to allow the jury to find that the raise was related to the travel contract, so we shall assume that this link has been established. The jury also learned that Craig Adelman, one of the principal owners and managers of Adelman Travel, supported Wisconsin’s Governor and made contributions to his campaign both before and after Adelman Travel was selected for this contract.

The prosecution does not contend, however, that any of these contributions was unlawful — they were properly disclosed, and no quid pro quo was entailed. There is not so much as a whiff of a kickback or any similar impropriety. Nor does the prosecution contend that Thompson knew or cared about these contributions. What, then, were the “political” considerations to which Thompson referred? We may assume that Thompson learned that her boss preferred Adelman Travel to Omega World Travel, and Thompson knew that Farley held a political rather than a civil-service appointment.

But why was Adelman Travel the favored bidder? One possibility is that Farley knew about, and sought to reward, Craig Adel-man’s past and potential financial support of the Governor. If that was Farley’s motive, then the selection was open to question under O’Hare Truck Service, Inc. v. Northlake, 518 U.S. 712, 116 S.Ct. 2353, 135 L.Ed.2d 874 (1996), and Board of County Commissioners v. Umbehr, 518 U.S. 668, 116 S.Ct. 2361, 135 L.Ed.2d 843 (1996), which hold that the first amendment limits the extent to which political support of office holders may justify the withholding of public contracts.

But these decisions do not say that the Constitution forbids all politically motivated contracting practices — and they certainly do not hold that any error in implementing the Supreme Court’s multi-factor-balancing approach is a crime. Perhaps, however, Farley favored Adel-man Travel because it was cheaper. This would be a political position in the best sense of that term. Many a person runs for office on a platform of cutting the cost of government.

Bureaucrats may call a preference for low price over high levels of service a form of “political interference” with their operations (especially when it is state employees who may suffer inconvenience in order to save the taxpayers’ money), but no party has a monopoly on opposing gold-plated wastebaskets and other excesses. Low prices may advance the public interest even if they discomfit public employees, and recognition that driving down the cost of government is good politics for incumbents does not transgress any federal statute of which we are aware.

Still another possibility is that Farley (and thus Thompson) sought to favor a local firm over one from another state. The Supreme Court has held that states, as market participants, may buy preferentially from their own citizens. See, e.g., White v. Massachusetts Council of Construction Employers, Inc., 460 U.S. 204, 103 S.Ct. 1042, 75 L.Ed.2d 1 (1983); Reeves, Inc. v. Stake, 447 U.S. 429, 100 S.Ct. 2271, 65 L.Ed.2d 244 (1980). A preference for in-state suppliers who can vote, over competitors who can’t, may be smart politics.

Again no federal statute regulates such behavior, let alone declares it to be a felony. Wisconsin law specifies a preference for domestic bidders, though only when the out-of-state bidder hales from a jurisdiction that favors its own citizens in procuremént decisions. Wis. Stat. § 16.75(l)(a) ¶ 2. ■ The evidence of record would not permit a jury to find beyond a reasonable doubt which of these three “political” reasons was Farley’s, let alone whether Farley’s reason also was Thompson’s — for Thompson may have been trying to be a faithful subordinate without questioning her boss’s bona fides.

Nor was the jury asked to determine Thompson’s motive. The United States maintains that Thompson’s objective is irrelevant. It is enough, the prosecutor insists, that Thompson deflected the decision from the one that should have been made under the administrative process. When coupled with a personal benefit (the raise), such a deflection is criminal under federal law, the United States insists.

In other words, the prosecutor’s argument is that any public employee’s knowing deviation from state procurement rules is a federal felony, no matter why the employee chose to bend the rules, as long as the employee gains in the process. (In stating the argument this way, we are assuming that the jury could and did find beyond a reasonable doubt that Thompson knew that the state’s procurement rules entitled Omega World Travel to the contract, given her fellow employees’ favorable view of Omega’s oral presentation.) Thompson was convicted of violating two federal statutes, and we start with § 666.

This statute applies to entities that receive more than $10,000 annually from the federal government, as Wisconsin does. (The state’s annual travel budget alone includes about $18 million in federal contributions.) Under this statute, (

a) Whoever ... (1) being an agent of an organization, or of a State, local, or Indian tribal government, or any agency thereof— , (

A) embezzles, steals, obtains by fraud, or otherwise without authority knowingly converts to the use of any person other than the rightful owner or intentionally misapplies, property that— (

i) is valued at $5,000 or more, and (ii) is owned by, or is under the care, custody, or control of such organization, government, or agency commits a felony. The prosecution’s theory is that Thompson “intentionally misap-plie[d]” more than $5,000 by diverting it from Omega World Travel to Adelman Travel.

This assumes, however, that a mistake is the same thing as a misapplication, and the statute does not say this. (It isn’t even clear that Thompson made a mistake: She had authority to order the best-and-final procedure, a statistical tie ensued, and the tiebreakers were carried out accurately.) Approving a payment for goods or services not supplied would be a misapplication, but hiring the low bidder does not sound like “misapplication” of funds. The federal government saved money because of Thompson’s decisions.

Section 666 is captioned “Theft or bribery concerning programs receiving Federal funds”, and the Supreme Court refers to it as an anti-bribery rule. See Sabri v. United States, 541 U.S. 600, 124 S.Ct. 1941, 158 L.Ed.2d 891 (2004); Fischer v. United States, 529 U.S. 667, 120 S.Ct. 1780, 146 L.Ed.2d 707 (2000); Salinas v. United States, 522 U.S. 52, 118 S.Ct. 469, 139 L.Ed.2d 352 (1997). Neither Thompson nor anyone else in state government was accused of taking a bribe or receiving a kickback. A statute’s caption does not override its text, but the word “misapplies” is not a defined term.

We could read that word broadly, so that it means any disbursement that would not have occurred had all state laws been enforced without any political considerations. Or we could read it narrowly, so that it means a disbursement in exchange for services not rendered (as with ghost workers), or to suppliers that would not have received any contract but for bribes, or for services that were overpriced (to cover the cost of baksheesh), or for shoddy goods at the price prevailing for high-quality goods. All of these conditions were satisfied in cases such as United States v.

Spano, 421 F.3d 599 (7th Cir.2005), and United States v. Martin, 195 F.3d 961 (7th Cir.1999). None is satisfied here. Faced with a choice between a broad reading that turns all (or a goodly fraction of) state-law errors or political considerations in state procurement into federal crimes, and a narrow reading that limits § 666 to theft, extortion, bribery, and similarly corrupt acts, a court properly uses the statute’s caption for guidance.

That plus the Rule of Lenity, which insists that ambiguity in criminal legislation be read against the prosecutor, lest the judiciary create, in common-law fashion, offenses that have never received legislative approbation, and about which adequate notice has not been given to those who might be ensnared. See, e.g., Staples v. United States, 511 U.S. 600, 619 n. 17, 114 S.Ct. 1793, 128 L.Ed.2d 608 (1994). Imagine how the prosecutor’s reading of § 666 would apply to a state official charged with implementing the Medicaid program.

Someone applies for payment of medical expenses; a state employee approves; later it comes to light that the applicant made just a little too much money to be eligible, so the decision was erroneous. A violation of regulations and perhaps of some statutes has occurred, but is the error a crime? As we read § 666, the answer is no unless the public employee is on the take or the applicant is a relative (for indirect benefits are another form of payoff).

An error — even a deliberate one, in which the employee winks at the rules in order to help out someone he believes deserving but barely over the eligibility threshold — is a civil rather than a criminal transgression. Likewise the sin is civil (if it is any wrong at all) when a public employee manipulates the rules, as Thompson did, to save the state money or favor a home-state producer that supports elected officials. Public employees often implement rules with which they disagree, and they are tempted to bend these rules to achieve what they deem better outcomes.

As long as the state gets what it contracts for, at the market price, no funds have been misapplied, even if the state’s rules should have led it to buy something more expensive (and perhaps of higher quality too). Evidence showing that Thompson believed that Adelman Travel provided less value for money than its competitors might support an inference that funds had been misapplied, but this record does not imply that Thompson’s deeds were at variance with her thoughts. Thompson was convicted under 18 U.S.C. § 1341 as well as under §

Section 1341 forbids “any scheme or artifice to defraud” that predictably employs the United States mails. What “fraud” did Thompson commit, and who was the victim? Thompson did not bilk the state out of any money or pocket any of the funds that were supposed to be used to buy travel. But in response to McNally v. United States, 483 U.S. 350, 107 S.Ct. 2875, 97 L.Ed.2d 292 (1987), Congress enlarged the scope of criminal fraud by enacting 18 U.S.C. § 1346.

This statute provides: For the purposes of this chapter, the term “scheme or artifice to defraud” includes a scheme or artifice to deprive another of the intangible right of honest services. The prosecutor’s theory, which the jury accepted, is that Thompson deprived Wisconsin of her “honest services” — that is, of her duty to implement state law the way the administrative code laid it down, with only 300 of 1,200 points apportioned according to price, while 200 points were available to the best-looking or most mellifluous oral presenter, even if Thompson deemed that allocation silly or counterproductive.

Once again that approach has the potential to turn violations of state rules into federal crimes. When the Supreme Court reverses a court of appeals, it is apt to say (as the prosecutor says about Thompson) that public officials have failed to implement the law correctly. Does it follow that judges who are reversed have deprived the United States of their honest services and thus committed mail fraud? (Judicial decisions are mailed.) How about the thousands of federal officials who contract for everything from pencils to aircraft carriers?

Disappointed would-be vendors may appeal through an elaborate bid-protest system — first to a Board of Contract Appeals, then to the Court of Federal Claims, and finally to the Federal Circuit. Reversals are common: Boards reverse contracting officers, the Court of Federal Claims reverses the Board only to be reversed in turn by the Federal Circuit. At each stage a reversal represents a conclusion that the predecessor has not implemented statutes and regulations correctly. But has anyone committed a crime in this sequence, unless he is on the take?

Because no one really thinks that § 1346 treats all legal errors by public employees as criminal, it has been necessary to cabin the reach of that statute. We held in United States v. Bloom, 149 F.3d 649, 655 (7th Cir.1998), that “[mjisuse of office (more broadly, misuse of position) for private gain is the line that separates run of the mill violations of state-law fiduciary duty ... from federal crimes.” The United States accepts this as the governing legal standard.

According to the prosecutor, Thompson “misused” her office when she lent it to political ends (even if the “political” end was just a lower price, about which incumbents could crow) and obtained a “private gain” when she got a raise. The prosecutor adds that, by currying favor with Farley, Thompson improved her job security. This is implausible; Thompson already had security as a civil servant. But cash is a form of gain. As we have explained, we proceed on the assumption that steering the contract to Adelman Travel is why Thompson received a raise.

Treating an incorrect application of state procurement law as a “misuse of office” and a raise as a “private gain” would land us back in the soup — once again, simple violations of administrative rules would become crimes. Nothing in the language of § 1341 or § 1346 suggests that Congress has created such ah equation, which would imply that every time a court sets aside a decision under the Administrative Procedures Act,'a crime has occurred if anyone involved in the administrative decision received a good performance review that led to a step increase under the General

Schedule of compensation. Suppose that a city’s mayor supports building a football stadium or power plant in Neighborhood A rather than B because residents of A want the project while residents of B have a NIMBY reaction, and the mayor concludes that he is more likely to be reelected if he can keep residents of both neighborhoods happy by ensuring that Neighborhood A gets the project. That is both self-interested (the mayor secures extra tenure) and public-interested (the populace as a whole is more satisfied).

That’s politics: people get and keep office by giving the people what they want, even if that is not what they “need” by some higher calculus. Imagine a governor who throws support (and public funding) behind coal-fired power plants because people fear nuclear power rather than because of a cost-benefit analysis; that may be a blunder but is not a crime even if the governor privately thinks that nuclear power would be superior. Similarly, supporting local vendors, a form of “pork,” may or may not be good government but is not a “misuse of office” in the criminal sense.

The idea that it is a federal crime for any official in state or local government to take account of political considerations when deciding how to spend public money is preposterous. “Misuse of office” may be almost as slippery a phrase as “honest sendees.” An error in carrying out one’s duties differs from “misuse” of position, but it is possible to imagine action as a result of impermissible motives that turns a mistake into a misuse of power.

This would be a hard case if the prosecutor had established that Thompson acted because of political contributions rather than, say, because she (and her boss) thought that buying services for less, or from local suppliers, is a form of good government from which incumbents would gain. One of these days we may need to gloss the phrase to reduce the risk that uncertainty poses to public servants. Today, however, it suffices to say that a raise approved through normal civil-service means is not the sort of “private gain” to which Bloom refers.

Section 1346 establishes that fraud committed against an employer need not put the employer out of pocket. To give one example, a state judge who accepts bribes has deprived the state of the service that it sought to purchase — adjudication that is impartial between the litigants. See United States v. Murphy, 768 F.2d 1518 (7th Cir.1985); United States v. Holzer, 816 F.2d 304 (7th Cir.1987). And the bribe produces private gain for the officeholder. All of the examples in Bloom involved payoffs outside proper channels.

Suppose, however, that the judge is wilful rather than corrupt — for example,'a judge in traffic court who dismisses speeding prosecutions in order to leave work early and play tennis on a sunny afternoon. Such a judge fails to provide the state with the adjudication service the populace wants and needs, but there is no side payment. There might be a personal gain beyond the tan; perhaps judges who resolve their cases with dispatch get promoted. This is often true with federal administrative law judges, who are more likely to receive step increases on the General

Schedule if they are current in their work. It would stretch the ordinary understanding of language, however, to call a public employee’s regular compensation, approved through above-board channels, a kind of “private gain.” Treating exercise or companionship on the tennis court as a private gain would be an even greater stretch, though leisure, friendship, and health all are substitutes for cash.

It is linguistically possible to understand “private gain” as whatever adds to the employee’s income or psyche — anything the employee would pay to have, rather than pay to avoid — but the Rule of Lenity counsels us not to read criminal statutes for everything they can be worth. The history of honest-services prosecutions is one in which the “private gain” comes from third parties who suborn the employee with side payments, often derived via kickbacks skimmed from a public contract.

Treating § 1346 as limited to such situations is consistent with its language: recall that it defines a means to implement a “scheme or artifice,” and getting a raise through normal personnel practices does not sound like an aspect of a “scheme or artifice.” The United States has not cited, and we have not found, any appellate decision holding that an increase in official salary, or a psychic benefit such as basking in a superior’s approbation (and. thinking one’s job more secure), is the sort of “private gain” that makes

an act criminal un-, der . § 1341 and § 1346.. The United States does rely on a few decisions of district courts, e.g., United States v. Sorich, 427 F.Supp.2d 820, 829 (N.D.Ill.2006); United States v. Munson, 2004 WL 1672880 *1, 2004 U.S. Dist. LEXIS 14274 *3 (N.D.Ill. July 27, 2004), but we do not find them persuasive. We now- hold that neither an increase in salary for doing what one’s superiors deem a good job, nor an addition to one’s peace of mind, is a “private benefit” for the purpose of § 1346.

Sections 666 and 1346 have an open-ended quality that makes it possible for prosecutors to believe, and public employees to deny, that a crime has occurred, and for both sides to act in good faith with support in the case law. Courts can curtail some effects of statutory ambiguity but cannot deal with the source.

This prosecution, which led to the conviction and imprisonment of a civil servant for conduct that, as far as this record shows, was designed to pursue the public interest as the employee understood it, may well induce Congress to take another look at the wisdom of enacting ambulatory criminal prohibitions. Haziness designed to avoid loopholes through which bad persons can wriggle can impose high costs on people the statute was not designed to catch. Thompson’s conviction is reversed, and the case is remanded with instructions to enter a judgment of acquittal.