Brotherhood Shipping Co., Ltd. v. St. Paul Fire & Marine Insurance Company and City of Milwaukee, and Afram Lines (International), Inc.

POSNER, Circuit Judge. The M/V Capetan Yiannis, a freighter, was damaged and temporarily put out of service in an accident in the Port of Milwaukee on Lake Michigan. The owner of the ship, Brotherhood Shipping Company, brought suit against the charterer (i.e., lessee) of the ship, Afram Lines, and also against the City of Milwaukee, which owns the port, and the city’s insurer, seeking damages for the damage to the ship and for the loss of revenues from the ship’s being out of service because of the accident. The plaintiff designated the suit as one in admiralty, in accordance with Rule 9(h) of the Federal Rules of Civil Procedure. The city counterclaimed for the damage that the accident had caused to the slip in which the ship was berthed at the time of the accident. Both the claim against the charterer and the counterclaim against the shipowner remain pending in the district court. The shipowner’s principal claim against the city (and its insurer — for Wisconsin is a direct-action state — but we can ignore the insurer), and the only one we need discuss, is that the city was negligent and its negligence contributed to the accident. The district judge granted the city’s motion for summary judgment and dismissed the shipowner’s claims against it. Since other claims remain pending in the district court, this was a partial dismissal only, making the shipowner’s appeal interlocutory. But as orders resolving the rights and liabilities of parties to admiralty *325 cases are appealable as of right even if interlocutory, 28 U.S.C. § 1292(a)(3), our jurisdiction is secure. The only issue for us is whether, viewing the evidence obtained through pretrial discovery as favorably to the shipowner as reason allows, we can say that no reasonable trier of fact could conclude that the city had been even a little bit negligent. It is important to emphasize that this is the only issue, because the appellees’ strategy in argument has been to direct our attention elsewhere. They argue that because this is a suit in admiralty and therefore to be tried by the judge rather than by a jury, we should permit the judge greater latitude in finding facts on summary judgment than we would if, were the case to be tried, the trier of fact would be different, would be a jury. And they argue that, even if the city was negligent, the captain of the Capetan Yiannis, and other persons who may (or may not) have been the shipowner’s agents, were themselves negligent — and far more so than the city. The first point is wrong, and the second irrelevant. Take the second first. The rule in admiralty, when property damage results from a collision between two ships, or, as here, between ship and shore, is comparative negligence (equivalently, relative fault). The plaintiff’s negligence reduces the amount of damages that he can collect, but is not a defense to liability. United States v. Reliable Transfer Co., 421 U.S. 397, 95 S.Ct. 1708, 44 L.Ed.2d 251 (1975). Now in many states that embrace comparative negligence the plaintiff’s negligence remains a defense to liability if he fails to prove that the defendant’s negligence was more than 50 percent responsible for the accident. Wisconsin is such a state. Wis.Stat. § 895.045. And the appellees urge us to apply Wisconsin’s rule. But admiralty is not merely a basis of jurisdiction, like diversity of citizenship. It is a body of substantive principles as well, or more precisely a grant of power to federal judges to create a body of admiralty law, much as section 301 of the Taft-Hartley Act, 29 U.S.C. § 185, has been interpreted as a grant not only of federal jurisdiction to enforce collective bargaining agreements but also of judicial authority to formulate federal rules of decision. Textile Workers Union v. Lincoln Mills, 353 U.S. 448, 456, 77 S.Ct. 912, 918, 1 L.Ed.2d 972 (1957). One of the judge-made substantive admiralty rules is the doctrine of comparative negligence, which in its original form as divided damages was invented in admiralty and applied there long before it became a part of the common law of torts through common law evolution and statutory intervention. (See also Pope & Talbot, Inc. v. Hawn, 346 U.S. 406, 408-09, 74 S.Ct. 202, 204, 98 L.Ed. 143 (1953), applying comparative negligence in personal-injury suits brought in admiralty long before most states had adopted comparative negligence.) The admiralty doctrine of comparative negligence, unlike some state comparative-negligence statutes, has never had a threshold above which a plaintiff’s negligence barred all recovery. We do not understand the city to be arguing that we should create such a threshold. It argues, erroneously, that we should apply Wisconsin law, which contains the threshold; but it does not argue that if we reject Wisconsin law we should create a new doctrine of admiralty law. As we are given no reason to depart from the “pure” comparative negligence standard (illustrated in admiralty by our decision in Cement Division v. City of Milwaukee, 915 F.2d 1154 (7th Cir.1990), and outside of admiralty by many cases, such as Lamborn v. Philips Pacific Chemical Co., 89 Wash.2d 701, 705, 575 P.2d 215, 219 (1978) (en banc); Villa v. Crown Cork & Seal Co., 202 Ill.App.3d 1082, 1084, 148 Ill.Dec. 372, 374, 560 N.E.2d 969, 971 (1990), and our Wassell v. Adams, 865 F.2d 849, 852 (7th Cir.1989)), we shall assume that if the plaintiff’s negligence is deemed 99 percent responsible for the accident in this case and the defendant’s negligence 1 percent responsible, the plaintiff is entitled to 1 percent of its damages. Of course, if the plaintiff’s negligence is such that the accident would have occurred even if the defendant had not been negligent at all, the plaintiff cannot recover, having failed to show a causal relation *326 between the defendant's negligence and the accident, an essential element in a tort suit. But it is not clear from the evidence gathered in discovery in this case that the accident would have happened even if the city had exercised due care; so we may disregard it as a possibility in this appeal. As for the fact that the case, if tried, would be tried by the judge who issued the order now under review by us granting summary judgment, it is not even clear that this is a fact. Setting aside every contingency that may cause a different judge to preside at a trial from the one who ruled on a motion for summary judgment, we point out that even a case within the admiralty jurisdiction is sometimes triable to a jury. Under the “saving to suitors” clause in the statute that specifies the admiralty jurisdiction of the federal courts, 28 U.S.C. § 1333(1), an admiralty plaintiff is entitled to pursue any other remedies that he might have as well as his remedies in admiralty. In this case, the requisites for diversity jurisdiction appear to be satisfied, and the plaintiff demanded a jury trial, to which the Seventh Amendment would entitle it in a diversity case in which damages are sought. While the plaintiff might not get far under Wisconsin tort law, for the reason already suggested (the difference between Wisconsin and admiralty comparative-negligence law), Wisconsin tort law would not apply. Admiralty law, and hence the admiralty rule on comparative negligence, furnishes the rule of decision in any case within the admiralty jurisdiction, even if it is within the diversity jurisdiction as well so that the plaintiff is entitled to demand a jury. Pope & Talbot, Inc. v. Hawn, supra, 346 U.S. at 410-11, 74 S.Ct. at 205-06; Harville v. Johns-Manville Products Corp., 731 F.2d 775, 779 (11th Cir.1984). What is true is that by pleading its claim as one in admiralty—for reasons, incidentally, that are wholly unclear—the plaintiff may well have waived the right to a jury trial that it would have had if it had forgone a Rule 9(h) designation and based jurisdiction instead on diversity of citizenship. T. N.T. Marine Service, Inc. v. Weaver Shipyards & Dry Docks, Inc., 702 F.2d 585 (5th Cir.1983) (per curiam); Romero v. Bethlehem Steel Corp., 515 F.2d 1249, 1253 (5th Cir.1975); 9 Charles Alan Wright & Arthur R. Miller, Federal Practice and Procedure § 2315 at p. 76 (1971); see also Note, “The Jury on the Quarterdeck: The Effect of Pleading Admiralty Jurisdiction When a Proceeding Turns Hybrid,” 63 Texas L.Rev. 533, 539-40 (1984). The plaintiffs decision to plead its case as an admiralty case cannot be dismissed as a formality, when already the plaintiff has used the distinctive procedures of admiralty to take this interlocutory appeal. We doubt profoundly that in these circumstances the district judge would—or should—be receptive to a motion under Rule 15 to amend the complaint by withdrawing the plaintiffs Rule 9(h) designation (though Rule 9(h) authorizes such motions) so that the plaintiff can have the best of both procedural worlds. But that is a side issue. The pertinent point is that there can be no certainty, even in a case within the admiralty jurisdiction, that the judge who grants summary judgment will, if his decision is reversed, be the trier of fact at trial. And in any event the care with which we review a grant of summary judgment ought not vary with the likelihood that if we reverse, the judge we are reversing will be the trier of fact on remand and will use his factfind-ing discretion to reinstate the judgment that we have reversed. Even realism has its limitations. It is no doubt inevitable that a reviewing court will give more or less weight to a district judge’s discretionary determinations depending on the respect in which the court holds the particular judge. It is neither inevitable nor desirable that the reviewing court should give more weight to a grant of summary judgment because the plaintiff has not asked for, or is not entitled to, a trial by jury if he clears the hurdle of summary judgment. For one thing such a differentiated standard of review would encourage plaintiffs to demand a jury trial where otherwise they might have been content with a bench trial. For another thing it would require us to get entangled in subtle issues, well *327 illustrated by this case, concerning the right to a jury trial. From time to time courts have hinted, though none to our knowledge has held, that if no jury trial is in the offing the trial judge can go beyond the strict scope of summary judgment — an inquiry into the existence of a genuine issue of material fact — and do a little factfinding. Houston North Hospital Properties v. Telco Leasing, Inc., 680 F.2d 19, 22, on rehearing, 688 F.2d 408 (5th Cir.1982); Tripp v. May, 189 F.2d 198, 200 (7th Cir.1951). Today we reaffirm that he may not. United States v. Bosurgi, 530 F.2d 1105, 1110, 1112 (2d Cir.1976); 10A Charles Alan Wright, Arthur R. Miller & Mary Kay Kane, Federal Practice and Procedure § 2728 at pp. 178-87 (1983). Of course if trial is to the court and the judge elects to proceed under Fed. R.Civ.P. 42(b) and try a discrete issue separately, that abbreviated “trial” may closely resemble a summary judgment proceeding, especially since a judge has a limited power to hear oral testimony in aid of determining whether there are any issues of fact that would prevent the grant of summary judgment. (Both points are discussed in Stewart v. RCA Corp., 790 F.2d 624, 628-29 (7th Cir.1986).) That is not this case. And litigants can, if they want, waive the right to a trial and treat the summary judgment proceeding as the trial, May v. Evansville-Vanderburgh School Corp., 787 F.2d 1105, 1115 (7th Cir.1986); Market Street Associates Limited Partnership v. Frey, 941 F.2d 588, 590 (7th Cir.1991), and then the judge should make factfindings just as he would in a real trial. That is not this case either. So we come back to where we began: The sole issue for us is whether, viewed as favorably to the plaintiff as is reasonable, the evidence gathered in pretrial discovery would, were the same evidence to constitute the entire record at trial, warrant a finding that the city was negligent. In answering that question, we apply the standard of negligence laid down by Judge Hand in the famous admiralty case of United States v. Carroll Towing Co., 159 F.2d 169, 173 (2d Cir.1947), endorsed by this court as the proper admiralty standard in United States Fidelity & Guaranty Co. v. Jadranska Slobodna Plovidba, 683 F.2d 1022 (7th Cir.1982), and cited approvingly by other courts in admiralty cases, as well. E.g., In re Complaint of Paducah Towing Co., 692 F.2d 412, 422 n. 18 (6th Cir.1982); Louisiana ex rel. Guste v. M/V Testbank, 564 F.Supp. 729, 740-41 (E.D.La.1983); Pruitt v. Allied Chemical Corp., 523 F.Supp. 975, 978 n. 11 (E.D.Va.1981); Selame Associates, Inc. v. Holiday Inns, Inc., 451 F.Supp. 412, 419 (D.Mass.1978); cf. East River S.S. Corp. v. Transamerica Delaval Inc., 476 U.S. 858, 872 (1986); Andros Shipping Co. v. Panama Canal Co., 298 F.2d 720, 726 (5th Cir.1962). Under that standard, a defendant is negligent if the burden (cost) of the precautions that he could have taken to avoid the accident (B in Hand’s formula) is less than the loss that the accident could reasonably be anticipated to cause (L), discounted (i.e., multiplied) by the probability that the accident would occur unless the precautions were taken. So: B<PL. The cost-justified level of precaution (B) — the level that the defendant must come up to on penalty of being found to have violated his duty of due care if he does not — is thus higher, the likelier the accident that the precaution would have prevented was to occur (P) and the greater the loss that the accident was likely to inflict if it did occur (L). Looked at from a different direction, the formula shows that the cheaper it is to prevent the accident (low B), the more likely prevention is to be cost-justified and the failure to prevent therefore negligent. Negligence is especially likely to be found if B is low and both P and L (and therefore PL, the expected accident cost) are high. We now have a framework for analyzing the facts. The Port of Milwaukee has a breakwater, with, of course, a gap in it to allow ships to go in and out of the harbor. Two of the slips in the harbor, where ships berth, are directly opposite the gap in the breakwater. As a result of this geometry, and the occasional violence of Lake Michigan storms, surprising to strangers to the Great Lakes, a northeast gale — and such gales are common on Lake Michigan — can *328 afflict these slips with two severe patterns of wave action, which the parties refer to as “cross-slapping” and “over-topping.” Waves from the open water surge through the gap in the breakwater, strike the wall on the other side, recoil, and collide with the next surge, creating a wave twice as high as the other waves stirred up in the harbor by the gale. This is cross-slapping. Over-topping occurs when waves in this area of the harbor become so violent that they spill over the wall of the slip at which a ship is berthed. The concern with these unusual conditions is not an academic one. Between 1964 and 1979, nine ships were damaged while berthed in the outer harbor (that is, in the area of the breakwater— there are safer births in an inner harbor). One of them sank. That disaster occurred in 1979, and the ship that sank had been berthed in the slip next to that in which the Capetan Yiannis became the tenth victim of cross-slapping and over-topping, in 1987. As early as 1951 the city had commissioned a study of wave conditions in the outer harbor that found those conditions to be unsafe. A series of additional studies, the last shortly before the accident that damaged the Capetan Yiannis, confirmed the hazard and recommended measures, such as the construction of a baffle device, that would reduce the violence of the waves in the outer harbor. The city did nothing. Well, not quite nothing. After the E.M. Ford sank in 1979 (precipitating the lawsuit that produced our decision in the Cement Division case, cited earlier), the city had the following notice inserted in the U.S. Coastal Pilot, the manual that mariners use in unfamiliar ports: “vessels moored in the outer harbor [of the Port of Milwaukee] may be subject to severe surging when there are strong NNE to ENE winds.” The city continued to promote the Port as a “harbor of refuge,” adding that the Port provides “everything you’d expect a major world port to have,” though in fact tugs and pilots are not available around the clock, as they are, at least according to some of the evidence, at other major ports. This evidence suggests, as we are about to see, that the master of the plaintiff’s ship may have been induced to rely on nonexistent safety features of the Port of Milwaukee. In December 1987, the Capetan Yian-nis, a 590-foot ocean-going freighter owned by a Greek company and captained by a Greek mariner named Konstadinos, was ordered to the Port of Milwaukee to pick up a cargo for South Africa. Konsta-dinos berthed the ship, pursuant to directions by a lessee that a trier of fact might find was the city’s agent, in one of the two slips opposite the gap in the breakwater. At noon the following day, while the ship was being loaded, Konstadinos heard (or was told about) a report on the radio that a gusty northeaster was brewing. He made some inquiries — not as yet from the city, however — and was reassured that he could ride out the storm. At 4 p.m., an hour after hearing a weather report similar to the one at noon, the city’s harbor master received a Coast Guard report predicting waves five to twelve feet high and winds of 35 to 40 knots out of the northeast. He wrote this information, together with the expected duration of the storm (36 hours), at the bottom of a “weather notice.” This is a printed form that states in its entirety: The Weather Bureau has warned the Milwaukee area that a North and/or Northeast storm is imminent. The resultant wave action can be very severe, causing your vessel to pitch, roll and yaw. These actions may be separate or simultaneous. This may result in damage to your vessel and/or the dock structure. This is not an order to vacate the berth. This is information to assist you in determining your course of action. The harbor master waited till he left for home at 5 p.m. to deliver the weather notice to Konstadinos, who, receiving it at 5:30 and becoming alarmed, sent for a pilot to direct the ship to a safer berth. The pilot, however, had — despite the stormy weather, which might have been expected to alert him to a possible emergency need for his services — already left for the day and was en route to his home in Chicago. Even if Konstadinos had wanted to take his *329 chances on leaving the berth without a pilot, he couldn’t have done so, because there were no tugs or linesmen available at that hour. By the time the pilot arrived, at 11:30 p.m., the storm was too severe for the ship to leave the berth. Conditions continued to worsen and at 6:15 a.m. the stern ropes broke and the stern of the ship was dashed against the wall of the slip. The plaintiff estimates its total damages— the cost of repairs plus lost revenue while the ship was out of commission — at $4.5 to $5 million. The owner of the cargo, whose claim has yet to be adjudicated by the district court, also sustained a loss. Evaluating these facts with the aid of the Hand formula, we note first that L in the formula — the magnitude or gravity of the loss that an accident that the precautions the defendant failed to take would have averted could be expected to inflict— was substantial. The ships that dock at the Port of Milwaukee are expensive machines carrying expensive goods. Moreover, an accident to such a ship, even while the ship is berthed, could endanger human life. The loss to the shipowner in this case has merely been alleged, not as yet determined, but there is nothing implausible in the claim that it is in the millions. The ship could have sunk, like the E.M. Ford, in which event the loss might easily have exceeded $5 million. (The damages sought for that sinking were $6.5 million. Cement Division v. City of Milwaukee, supra, 915 F.2d at 1156.) As for the likelihood of such an accident (P), it could not be reckoned small, given the history of accidents to ships at the two exposed slips in the outer harbor. So PL, the expected accident cost, was substantial and therefore imposed on the defendant a duty of taking substantial precautions. At least three types of precaution were possible. One was a structural alteration to the harbor that would have eliminated the problem once and for all, or, more realistically, have greatly reduced the risk of accident. The second possible precaution was to have pilots, tugs, and linesmen available round the clock in the winter to remove endangered ships from the slips at short notice. The third was to give the masters of the ships berthed at the “bad” slips (a characterization offered by one of the city’s own employees) sufficient early warning to enable them to obtain a pilot, tugs, and linesmen before the close of business on the day of a storm. The city did none of these things. It did not build a baffle or other device that would have prevented cross-slapping and over-topping. It did not arrange for pilots, tugs, and linesmen to stand by after 5 p.m. during the winter in case there was a storm. It did not include an explicit warning in the U.S. Coastal Pilot —the note on weather conditions makes no reference to the bad slips. It did not direct the Capetan Yiannis to an empty slip (if there was one) in a less exposed position. And it made no effort to communicate a clear and effective warning to Captain Konstadinos as soon as the harbor master found out that a dangerous northeaster was approaching. Maybe every one of these precautions would have cost more than the benefit in averting the accident. But on that subject all that the city says is that it doesn’t employ any pilots — they are independent contractors who deal directly with the ship’s crew. This is hardly responsive. It is not beyond the city’s power to make an arrangement with the pilots’ association under which the association would undertake to have a pilot on stand-by status after 5 p.m. But this is not an issue worth pursuing here. Clearly the cheapest precaution was to include a candid, comprehensive explanation in U.S. Coastal Pilot. Second cheapest was to give a prompt warning if and when a storm was expected. In some cases the warning would come too late. But here the city’s harbor master knew by 3 p.m. and probably earlier that a northeaster was approaching, and when he received the Coast Guard report at 4 p.m. he knew everything a reasonable person would have had to know to have a duty to warn the master of the Capetan Yiannis immediately that he must make arrangements to clear the slip at short notice. The printed weather notice was inexplicit but in any event was delivered too late. *330 Given the unpredictability of Lake Michigan storms and the vagaries of a notice system, sole reliance on such a system to protect mariners, many of whom like Captain Konstadinos are foreigners unfamiliar with Lake Michigan, might not suffice to fulfill the city’s duty of care, provided one of the alternatives — a structural alteration to the harbor or the provision of stand-by rescue services — would be less costly than the expected accident cost that would remain once the notice system was in place. The U.S. Coastal Pilot warning, even if far more explicit than it was, might have been deemed inadequate if, in conjunction with the city’s promotional literature, it left the impression that adequate rescue services were available on a stand-by basis, which they were not. But we need not get into any of these issues. They may or may not become important on remand. For purposes of this appeal it is enough that the plaintiff has raised a genuine issue of material fact concerning the adequacy of the precautions that the City of Milwaukee took to prevent the type of serious, and by no means remotely unlikely, accident that occurred. In so saying we do not deny the possibility that the accident might have been prevented had Captain Konstadinos been more alert, more cautious, or spared the reassuring advice of others, including a charterer’s agent who may be an agent of the plaintiff’s. Or perhaps the captain should have been more daring, and have left the berth without waiting for a pilot and tugs and linesmen. But to repeat an earlier point, the issue of the parties’ relative fault is not before us. There is evidence from which a reasonable trier of fact could infer that the city was negligent and that the accident would not have occurred had it not been. No more need be decided to require that the judgment of the district court be REVERSED.