The United States appeals from an order by the district court (reported at 634 F.Supp. 435 (N.D.Ill.1986)) allowing Amexchem — which had imported certain new animal drugs into the United States in alleged violation of the Federal Food, Drug, and Cosmetic Act, 21 U.S.C. §§ 301 et seq. — to re-export the drugs. The Food and Drug Administration had seized them pursuant to 21 U.S.C. § 334, and wanted to destroy them. Because the order allowing the drugs to be re-exported wasn’t stayed, Amexchem went ahead and re-exported them before the appeal was heard. In its brief the United States urged us to dismiss the appeal (its appeal) from the order allowing re-exportation as moot — since, right or wrong, the drugs had been re-exported— and to vacate the district court’s decision and order, which we must do when a district court decision becomes moot pending appeal, see Great Western Sugar Co. v. Nelson, 442 U.S. 92, 99 S.Ct. 2149, 60 L.Ed.2d 735 (1979) (per curiam); United States v. Munsingwear, Inc., 340 U.S. 36, 71 S.Ct. 104, 95 L.Ed. 36 (1950); Gjertsen v. Board of Election Comm’rs, 751 F.2d 199, 202 (7th Cir.1984). At argument, however, the government changed its mind and urged us to reach the merits and reverse. We think it was right the first time.

The case involves 12 lots of new animal drugs in bulk form which Amexchem imported into the United States in 1983 and 1984 and which the FDA seized in 1984 from the warehouses in which Amexchem had stored them pending sale to American companies. The ground of the seizure was that the drugs were misbranded, adulterated, or both. These are words of art under the statute. Ten of the lots were “misbranded” because the directions for use on their labels were inadequate and three of these also because the labels were in a foreign language. See 21 U.S.C. § 352. All twelve of the lots were “adulterated” (in the FDA’s view) because Amexchem did not have a firm order from firms licensed to use the drugs. See 21 U.S.C. §§ 351(a)(5), 360b. There is no dispute about the misbranding, so let us focus on the two lots that were adulterated but not misbranded. Both were lots of oxytetracycline. At least two companies in the United States have licenses to use oxytetracycline to make finished drugs that are mixed with animal feed, and these were the companies that Amexchem hoped to sell the oxytetracycline to. But the FDA takes the position that the statutory provisions that deem an animal drug adulterated “unless ... there is in effect an approval of an application ... with respect to ... such drug,” 21 U.S.C. §§ 351(a)(5), 360b(a)(l)(A), require that Amexchem, which did not itself have a license to use the drug, have firm orders from the licensees to whom it intended to sell it. Amexchem did not have any firm orders for the oxytetracycline; all it had was its hopes that licensees would buy the drug from it. The FDA’s concern is that if an importer such as Amexchem stores drugs in a warehouse for a substantial period of time while looking for an authorized purchaser, the FDA may lose track of them and they may be put to some unauthorized use that might endanger the safety of the animals fed the drugs and of the people who eat the meat of those animals.

When imported items are condemned under the food and drug act, the district court “may permit the article to be delivered to the owner for exportation in lieu of destruction,” provided that the importer “had no cause for believing that it was adulterated, misbranded, or in violation before it was released from customs custody.” 21 U.S.C. § 334(d)(1). The district court found that Amexchem had had no actual notice that the misbranded lots did not comply with the various misbranding regulations; for example, some of the misbranding resulted from manufacturers’ shipping drugs[*571] to Amexchem contrary to Amexchem’s orders. As for adulteration, the court found that Amexchem had neither actual nor constructive notice of the FDA’s “firm order” policy. The policy is not contained in any published regulation, and in the court’s view is not so straightforward an interpretation of the statute that it can lawfully be imposed on an importer without notice-and-comment rulemaking. Hence the court exercised its discretion to permit Amexchem to re-export the 12 lots (worth in the aggregate about a half million dollars) rather than suffer their destruction.

For purposes of this appeal the government does not contest the court’s findings with regard either to misbranding or to Amexchem’s lack of actual notice of the FDA’s “firm order” policy. But it strongly disagrees that it must undergo the delay and expense of notice-and-comment rule-making, or for that matter any other form of rulemaking, in order to effectuate the “firm order” policy. The FDA analogizes its position to that of the Labor Board when adopting a new interpretation of the National Labor Relations Act in an unfair labor practice proceeding, or that of the Federal Trade Commission when discovering a new form of unfair competition in an adjudicative proceeding under the Federal Trade Commission Act. Cf. Mosey Mfg. Co. v. NLRB, 701 F.2d 610, 612 (7th Cir. 1983) (en banc). It has withdrawn its suggestion of mootness because of the enormous impact that the district court’s decision could, if followed by other courts, have on the control of imports of new animal drugs, and because of the uncertainty into which the decision has plunged importers of those drugs.

This is, or at least began as, an in rem case, which is to say a case nominally against a thing rather than a person. The res, which consists of the drugs that the government seized but that Amexchem was later permitted to re-export, is no longer within the jurisdiction of the court. No res, no case. United States v. 3 Unlabeled 25-Pound Bags Dried Mushrooms, 157 F.2d 722 (7th Cir.1946). The government, however, claims to have personal jurisdiction over Amexchem, and Amexchem does not contest the claim. The action is therefore in personam as well as in rem. United States v. An Article of Drug Consisting of 4,680 Pails, 725 F.2d 976, 982-84 (5th Cir.1984). So the disappearance of the res does not resolve the issue of mootness.

That issue is complicated by a fundamental ambiguity in the district court’s opinion. The only issue before the court was whether to permit Amexchem to re-export the 12 lots — for having failed, in fact, to find customers for them, Amexchem did not want to continue holding them in its U.S. warehouses. The court's opinion could be read to mean that: the “firm order” policy is not sufficiently obvious for an importer to be charged with notice of it; therefore Amexchem had no cause to believe that merely because it had no firm orders for the drugs it couldn’t lawfully import them; therefore it can be permitted, under 21 U.S.C. § 334(d)(1), to re-export them. If this is all the court meant, the case is clearly moot. Whether or not Amexchem had actual notice of the “firm order” policy back in 1983 when it imported these drugs, it has such notice now. So it could not now import additional drugs for which it had no firm orders, and hope to stave off their destruction; therefore, overturning the district court’s decision would not affect Amexchem’s behavior. Granted, the FDA might gain something from a reversal of the decision — might gain a useful precedent to brandish in disputes with other importers. But that kind of advantage cannot prevent a case from becoming moot. No federal court may issue an opinion that, while it might provide guidance for the resolution of disputes not before the court, would not resolve an existing, living dispute before the court, either because there is no such dispute or because the dispute has been resolved by other means. If, by re-exporting the drugs in circumstances where it cannot possibly re-import them (or import other drugs) without inviting their destruction, Amexchem has ended its dispute with the FDA, any decision we rendered would be just such an “advisory” opinion, useful only in resolving disputes that have not yet become the subject of a lawsuit.

[*572] We add, irrelevant though it may strictly be, that the FDA has little to fear from the precedential effect of a district court decision vacated because the dispute has become moot. It is true that vacating a decision because of supervening mootness does not destroy its precedential effect. The purpose of setting aside a decision on that ground is only to prevent the decision from having res judicata or collateral estoppel effect in future cases. See, e.g., United States v. Munsingwear, supra; Commodity Futures Trading Comm’n v. Board of Trade of City of Chicago, 701 F.2d 653, 656-57 (7th Cir. 1983). The district court’s decision, when made, was within the court’s power to make, because the case wasn’t moot then. Its later becoming moot prevents appellate review and by doing so deprives the decision of the additional precedential force that it would have had if it had been affirmed by a higher court, but it does not eliminate whatever precedential force an unreviewed and unreviewable lower-court decision has. A single district court decision, however (especially one that cannot be appealed), has little precedential effect. It is not binding on the circuit, or even on other district judges in the same district. Colby v. J.C. Penney Co., 811 F.2d 1119, 1124 (7th Cir.1987).

This case may seem to fit the exception to the conventional principles of mootness that keeps a case alive where it would otherwise “evade review.” As an original matter we might question whether even Roe v. Wade, 410 U.S. 113, 125, 93 S.Ct. 705, 712, 35 L.Ed.2d 147 (1973), presented a suitable case for relaxing the usual standards of mootness. It is true that a woman desiring an abortion could not complete a lawsuit challenging an anti-abortion law till it was too late for her to have the abortion, but doctors prosecuted for performing illegal abortions could be permitted to litigate the legality of the prosecution, at their leisure, by raising their constitutional defense in the criminal action. Still, the position of the doctor is different from that of the patient, and she could never litigate her case to the end if the conventional principles of mootness were followed. This case is different. A party in Amexchem’s position is quite likely to be able to litigate its claim to completion. Anytime such a firm is denied the privilege of re-exportation, or the government succeeds in getting a stay of an order permitting re-exportation, the case will not become moot, as it has become here as a result of the importer’s being allowed to re-export the drugs before the appeal could be decided. The pressure to relax the traditional rules of mootness that was felt in Roe v. Wade is thus not present here.

There is a more fundamental difference between Roe and this case. As explained in Weinstein v. Bradford, 423 U.S. 147, 149, 96 S.Ct. 347, 348, 46 L.Ed.2d 350 (1975) (per curiam), the principle of Roe applies only where there was a “reasonable expectation that the same complaining party would be subjected to the same action again.” (Emphasis added.) There is some probability that a woman who wants an abortion but is forbidden to get one will get pregnant again and want to abort the new foetus and again be unable to complete a legal challenge to the statute that forbids abortions. If, however, Amexchem, having now received (we are assuming) notice that it requires a firm order, is highly unlikely to import drugs without such an order, then the requirement of a “reasonable expectation” that it will be subjected to another action to seize and destroy its goods because it lacks a firm order is not satisfied; for without a firm order it will not import.

Amexchem points to ominous signals that the government is planning to prosecute it criminally for the alleged misbranding and adulteration. Generally, however, a civil court will not enjoin a criminal proceeding, see, e.g., Hynes v. Grimes Packing Co., 337 U.S. 86, 98-99, 69 S.Ct. 968, 976-77, 93 L.Ed. 1231 (1949); Ivy v. Katzenbach, 351 F.2d 32, 34-35 (7th Cir.1965), and Amexchem has made no effort to show that this case is within one of the exceptions to this rule. We think it follows that this civil suit cannot be kept alive merely because a favorable outcome would scotch a criminal prosecution. But[*573] cf. Williams v. Immigration & Naturalization Service, 795 F.2d 738, 741 (9th Cir. 1986). In any event, there is many a slip betwixt a criminal investigation and an actual conviction. It is pure surmise that Amexchem will be punished criminally, or that the course of any criminal proceeding would be affected by our deciding the merits of this appeal.

Thus far we have assumed that the district court’s opinion merely teases out the implications of “cause to believe” in the statute that authorizes the court to spare the importer’s goods from destruction. But the opinion can easily be read more broadly — indeed, this is the natural reading — to hold that the “firm order” policy is invalid and will remain so until and unless the FDA promulgates it as a formal rule, which the FDA will not do unless we uphold the district court’s decision. If this is the correct interpretation of the district court’s decision, and if when so interpreted the decision is correct, it follows that until the FDA gets around to promulgating a formal rule Amexchem can import new animal drugs for which it has no firm orders without being guilty of adulteration. Actual notice would be irrelevant, because the threshold requirement for destruction— that the imported goods have been properly condemned — could not be satisfied.

When the district court’s opinion is read so, the probability of Amexchem’s again importing new animal drugs without firm orders for them rises. No longer can one be confident that this district judge, at any rate, would deny the privilege of re-exportation just because Amexchem could not, the second time around, argue that it still lacked actual notice of the FDA’s “firm order” policy; the judge thinks the policy invalid. But the question then becomes, how probable is it that Amexchem in these circumstances would re-import these drugs, or import other drugs, without firm orders? If it is highly improbable, the case is moot.

As so often in law, deciding whether a case is moot or not moot requires a dichotomous solution to a continuous problem. See Commodity Futures Trading Comm’n v. Board of Trade of City of Chicago, supra, 701 F.2d at 655. We can never be certain that our decision of an appeal will have any effect, other than the insufficient one of providing guidance for future cases. For example, we might reverse an order enjoining the construction of a nuclear power plant only to discover that the defendant had decided to abandon the project anyway. Similarly, we might decide that the FDA’s “firm order” policy cannot be effectuated without a rule-making proceeding only to discover that Amexchem for a variety of reasons won’t ever again import new animal drugs into the United States without a firm order. We think the probability of this result is so high as to make the case moot. Amexchem’s last experience with importing such drugs without a firm order was not a happy one, quite apart from the legal troubles it got into. Evidently it could find no American customers for the drugs, and was happy to re-export them rather than to defend in this court the district judge’s (apparent) ruling that it was entitled to import them without firm orders. Most important, although there has been an issue of potential mootness ever since Amexchem asked the district judge to allow it to re-export the drugs rather than suffer their destruction, the record contains no evidence that Amexchem has any plans to import more drugs into the United States— with or without firm orders. It is not a large company, and it appears not to have any regular business of importing animal drugs into this country. Those 12 lots (plus a few imported in 1982) may have been an experiment, and an unhappy one. Amexchem has imported no drugs since 1984, and while the FDA’s seizure may be the reason, that is just a conjecture.

It is true that United States v. Generix Drug Corp., 460 U.S. 453, 456 n. 6, 103 S.Ct. 1298, 1300 n. 6, 75 L.Ed.2d 198 (1983), relying on such earlier cases as United States v. W.T. Grant Co., 345 U.S. 629, 632, 73 S.Ct. 894, 897, 97 L.Ed. 1303 (1953), states flatly that “the possibility that respondent may change his mind in the future is sufficient to preclude a finding of mootness.” Generix, however, and all the[*574] other cases in its line, are cases where the government had obtained an injunction against some practice and the defendant argued that because it had discontinued the enjoined activity the case was moot and should be dismissed, injunction and all. In Generic, for example, the court of appeals had vacated the injunction and the government had appealed to the Supreme Court; it was at this point that the defendant (respondent in the Supreme Court) tried to get the case dismissed as moot, in order to protect its victory in the court of appeals. The present case is different. There is no injunction. The only thing the government is seeking is an order to destroy drugs which, since they have been re-exported, are no longer within the power of the government to destroy. The only benefit the government can get from reversing the district court’s decision is the benefit of a favorable precedent which may affect Amexchem’s future plans — in other words, the benefit of an advisory opinion favorable to the government’s position.

Since the drugs were not re-exported till the district judge denied the government’s motion to stay his order permitting re-exportation, it may seem that neither party has had a chance to show that the case really is not moot. But as we have said, once the motion was made, and especially after it was granted, the parties knew there was a potential problem of mootness. If they wanted to keep the case alive (as they do, Amexchem in the hope of warding off a criminal prosecution, the government in the hope of eliminating an adverse precedent), it behooved them to present evidence, whether by supplementary proceedings in the district court or in this court, showing that the government has a live controversy with Amexchem even though the drugs that were the immediate subject of the dispute between the parties are back where Amexchem wanted them to be. The government presented no evidence; nor did Amexchem, despite its desire to avoid a finding that the case has become moot.

If we thought it highly probable as a matter of common sense or common knowledge that Amexchem would resume importation of drugs to the U.S., even without firm orders for them, provided we held that it had the right to do so, we would not dismiss the case as moot, since as we have said a judgment of mootness is, realistically, a judgment of more or less rather than of yes or no. But so far as we can guess from the scanty materials in the record, the probability is small. The business risks are great and there is no indication that Amexchem is prepared to run them. Its insistence that we reach the merits may seem to belie this conclusion, but is explained as we have said by Amexchem’s desire to improve its position in the criminal investigation — a desire that cannot save a case from a finding that it is moot.

As a detail, we note that there is some question whether Amexchem’s president, Dali, is a party to this case. If not, his intentions regarding the future importing of drugs without firm orders are irrelevant; if he is, still there is no more evidence that he would take a chance on doing this than there is that Amexchem itself would.

The judgment of the district court is vacated with directions to dismiss the lawsuit and vacate all orders entered by the court. No costs will be awarded in this court.

Vacated as Moot.