Rexnord, a manufacturer of roller chain, hired DeWolff Boberg & Associates (DBA), a management-consulting firm, to help improve its productivity. The parties had a falling out and Rexnord brought this diversity suit for breach of contract, obtaining a jury verdict of some $1.6 million, exactly one-half of what it had sought. To that amount the judge added an award of attorneys’ fees of almost $600,000 pursuant to a term of the contract. Indiana law is agreed to govern all the substantive issues in the litigation.
The consulting agreement was drafted by DBA and provides that DBA “shall provide, on a timely basis, and to the best of its ability, the professional services as generally outlined in the Proposal from the Analysis Report.” Going to the Proposal section of DBA’s precontractual analysis of Rexnord’s needs (the “Analysis Report”), one discovers a whole series of ‘We wills”: “We will design and install management systems to give your supervisors and managers the information they need to effectively control all of the functions within their departments.” We will train your supervisors and managers ‘on the floor,’ so they truly understand how to apply and use the systems and management concepts in their operations.” We will conduct a series of opening meetings during the first two weeks, to set the stage for the process that is starting.” We will develop a performance improvement evaluation method to measure the attainment of our savings commitment on a weekly basis.” “We will design the supervisory training program based on the weaknesses we observed.” *1003“We will design the management operating system upgrades with the supervisors to provide all of the currently missing elements while fine tuning marginal elements [including] ... staffing requirements determination.” And so on.
The consulting agreement also states that
we [DBA] estimate the annual savings available to you through implementation of our program will be Four Million Dollars ($4,000,000). These savings will come about from increased labor productivity, increased throughput, and reduced rework and scrap.... Please note that we are not attempting to put a financial value on the many collateral benefits that will come about as a result of this program, such as improved customer service, improved employee morale and ongoing improvements made by your people using this process,
and adds that DBA’s fee is $1.32 million, payable weekly over the course of the project; but if DBA is not “able to achieve our guaranteed annual savings rate [the $4 million] by the planned completion date of our program [the program was to run for 33 weeks], we would either continue working on your premises at our expense, or we would reimburse a portion of our fees, to provide you with the original estimated return on investment.”
The 33 weeks ended in August of 1997, but (we are now summarizing the evidence developed at trial, construed as favorably to Rexnord as the record permits) the promised $4 million savings had not yet been achieved and DBA chose to continue working. Why had the savings not materialized? Well, DBA had recommended that Rexnord reduce the number of its workers in certain departments, and Rexnord had complied. But the recommendation had proved to be a bad one, so DBA had changed course and recommended that workers be moved into those departments from other departments. When this failed too, DBA had changed course once again and urged Rexnord to hire a number of new employees, which Rexnord did. These about-faces in personnel policy caused the morale of Rexnord’s employees to plummet, and in August more than half the employees could be seen wearing T-shirts emblazoned “DON’T MISMANAGE OUR JOBS AWAY,” with the letters D, B, and A emphasized. Turnover among both supervisors and workers had soared, causing impaired productivity and additional severance and recruitment expenses. The company lost business and market share, customers were permanently lost, profits fell, rework and scrap increased, overtime increased, delivery reliability deteriorated. All these were consequences of the disruptive effects of DBA’s inept recommendations regarding staffing; and the poor quality and defective implementation of the recommendations could be traced in turn to breaches by DBA of specific promises in the Proposal, such as the promise to install a management operating system that would determine optimal staffing requirements and the promise to train supervisors and employees in the new systems. (DBA objected to a number of Rexnord’s damages exhibits on the ground that Rex-nord had failed to prove a causal relation between breach and damages; but this just is not so.)
DBA gave up the project in October, still having achieved nowhere near the $4 million in promised savings; but it refused to refund any part of its fee, on the ground that the failure of the project had been due to lack of cooperation by Rexnord. DBA complains that the judge failed to instruct the jury adequately on this defense (that the promisee made it impossible for the promisor to carry out its promise); but he did, if less amply than would have been desirable, by stating that Rexnord could *1004not prevail without proving that it had “performed its duties under the contract.”
Contract law distinguishes between direct and consequential damages, the difference lying in the degree to which the damages are a foreseeable (that is, a highly probable) consequence of a breach. See Evra Corp. v. Swiss Bank Corp., 673 F.2d 951, 958 (7th Cir.1982); Suburban Propane v. Proctor Gas, Inc., 953 F.2d 780, 785 (2d Cir.1992); Vitol Trading S.A. v. SGS Control Services, Inc., 874 F.2d 76, 79 (2d Cir.1989); R.K. Chevrolet, Inc. v. Hayden, 253 Va. 50, 480 S.E.2d 477, 481 (1997); cf. AES Technology Systems, Inc. v. Coherent Radiation, 583 F.2d 933, 940-41 (7th Cir.1978). DBA’s refusal to refund any part of its fee when the project failed to generate the promised savings had the utterly foreseeable, indeed certain, consequence that Rexnord paid more for DBA’s services than it had agreed to do. That is an example of direct damages. In contrast, the effect on Rexnord of the breach of various undertakings by DBA was more difficult for DBA to foresee because it depended to a large extent on matters internal to Rexnord.
Contract law takes two approaches to consequential damages in cases in which the contract itself fails to make provision concerning them (this is such a case). One, which the great Holmes favored (see Globe Refining Co. v. Landa Cotton Oil Co., 190 U.S. 540, 543-45, 23 S.Ct. 754, 47 L.Ed. 1171 (1903)) but has fallen into disuse, e.g., Western Industries, Inc. v. Newcor Canada Ltd., 739 F.2d 1198, 1203 (7th Cir.1984); Native Alaskan Reclamation & Pest Control, Inc. v. United Bank Alaska, 685 P.2d 1211, 1219 (Alaska 1984), though not everywhere, see Stifft’s Jewelers v. Oliver, 284 Ark. 29, 678 S.W.2d 372, 373 (1984), is that consequential damages can be awarded only if the promisor has assumed the risk of the consequences in question — has, in other words, agreed, whether expressly or as a matter of “what the parties probably would have said if they had spoken about the matter,” to bear them. Globe Refining Co. v. Landa Cotton Oil Co., supra, 190 U.S. at 543, 23 S.Ct. 754; see also Hector Martinez & Co. v. Southern Pacific Transportation Co., 606 F.2d 106, 109 n. 6 (5th Cir.1979). DBA’s explicit promise of $4 million in savings was a promise of just that character. The second approach requires merely that consequential damages be foreseeable, Western Industries, Inc. v. Newcor Canada Ltd., supra, 739 F.2d at 1203, and it is the approach that the parties have correctly assumed governs here. Rogier v. American Testing & Engineering Corp., 734 N.E.2d 606, 614 (Ind.App.2000); Indiana Ins. Co. v. Plummer Power Mower & Tool Rental, Inc., 590 N.E.2d 1085, 1092 (Ind.App.1992); Schroeder v. Barth, Inc., 969 F.2d 421, 425 (7th Cir.1992) (Indiana law).
Were the consequential damages foreseeable here? The famous case of Hadley v. Baxendale, 9 Ex. 341, 156 Eng. Rep. 145 (1854), held that a mill could not recover lost profits as a result of a carrier’s delay (in breach of the contract of carriage) in delivering a part essential to the mill’s operation; and there is a superficial resemblance, supportive of DBA’s position, to this case. But the cases are different in two important respects. First, unlike the carrier in Hadley, DBA knew a great deal about the promisee’s business and should have been able to estimate the likelihood and magnitude of the losses that it would impose on a client if it failed to honor its undertaking. Second, the result in Hadley may have depended on the mill’s failure to have protected itself against the consequences of a delay by the carrier by having a spare part on hand. Rodi Yachts, Inc. v. National Marine, Inc., 984 F.2d 880, 884 (7th Cir.1993); Rardin v. T *1005 & D Machine Handling, Inc., 890 F.2d 24, 27 (7th Cir.1989); Evra Corp. v. Swiss Bank Corp., supra, 673 F.2d at 957; Coastal Power Int'l, Ltd. v. Transcontinental Capital Corp., 10 F.Supp.2d 345, 368 n. 166 (S.D.N.Y.1998), aff'd, 182 F.3d 163, 165 (2d Cir.1999); see also Wells Fargo Bank, N.A. v. United States, 33 Fed.Cl. 233, 242 (1995), aff'd. in part, rev’d in part, 88 F.3d 1012 (Fed.Cir.1996). In other words, there was a sense in which the mill was the author of its own loss. In still other words, the carrier was not on notice that the mill was taking an unusual risk and had it been told that it would be an insurer against that risk it would have demanded a higher price for its service. DBA argues that Rexnord was likewise the author of its own loss, but the evidence permitted the jury to reject the argument.
What is slightly offputting about the consequential damages sought and awarded in this case is that they seem rather more like tort damages than like contract damages. Suppose that in the course of performing the contract, which required DBA personnel to spend a lot of time in Rexnord’s factory, an employee of DBA had through carelessness dropped his cell phone into the assembly line, causing damage. Unless the contract expressly or implicitly required the exercise of due care by DBA personnel when at Rexnord’s plant, Rexnord’s remedy would he in tort rather than in contract even though the injury had arisen in the course of the performance of a contract. Raquet v. Thompson, 693 N.E.2d 969, 971 (Ind.App.1998); Crum v. AVCO Financial Services of Indianapolis, Inc., 552 N.E.2d 823, 827 (Ind.App.1990) (“the negligent performance of a contract may give rise to liability in tort”); Essex v. Ryan, 446 N.E.2d 368, 370-71 (Ind.App.1983) (“it is axiomatic that one who contracts to perform services may commit both a breach of contract and the tort of negligence when he negligently fails to perform in a workmanlike manner”). This case looks like that because one way to describe it is as a case in which DBA’s people came into Rexnord’s plant and messed things up albeit without causing physical damage to property. But the difference is that the mess-up resulted from a failure to comply with express terms of the contract; and when a standard of care is specified in the contract it supersedes the tort standard (for why shouldn’t the parties be able to cut their own deal, so far as their rights and liabilities to each other are concerned?) and the promisee’s remedy is limited to a suit for breach of contract. Wells v. Stone City Bank, 691 N.E.2d 1246, 1249 (Ind.App.1998).
So far, so good. But unfortunately the appeal, and indeed the entire litigation, have been muddied by the misplaced preoccupation of both parties with the “collateral benefits” provision of the consulting agreement, in which, the reader will recall, DBA stated that it was “not attempting to put a financial value on the many collateral benefits that will come about as a result of this program, such as improved customer service, improved employee morale and ongoing improvements made by your people using this process.” DBA argues that this provision did not create a legally enforceable duty but that the damages the jury awarded are based in part on the erroneous belief that it did. Rexnord insisted at trial and continues to insist that the provision indeed imposed a legally enforceable duty that DBA violated.
The provision is a red herring. Rexnord to the contrary notwithstanding, the provision does not contain a binding promise, does not justify any award of damages, and indeed has nothing to do with this case. The very term “collateral,” the placement of the provision after the express promise of $4 million in savings, and DBA’s disclaimer of being able to put a *1006price tag on the benefits, scotch any inference that a binding promise was intended. (Language purists, which we however are not, would wish to point out in addition that third-person commitment is expressed by “shall,” not “will,” the reverse being true for first-person commitment. “We will undertake” and “he shall undertake” express commitment — and notice that the promises in the Proposal are in the form “we will.” “We shall undertake” and “he will undertake” are statements merely of intention or prediction.)
Then too, DBA would be unlikely to make a binding promise to achieve a result so dependent on the conduct of the promis-ee as “ongoing improvements made by your people using this process.” Such a promise would place DBA at Rexnord’s mercy, cf. Charter Oak Fire Ins. Co. v. Color Converting Industries Co., 45 F.3d 1170, 1174 (7th Cir.1995); Wallace v. Employers Mutual Casualty Co., 443 So.2d 843, 849 (Miss.1983), with only the limited protection provided by the doctrine of mitigation of damages. Outboard Marine Corp. v. Babcock Industries, Inc., 106 F.3d 182, 185 (7th Cir.1997).
The clincher is the sheer impossibility of quantifying the damages that Rexnord incurred from DBA’s failure to achieve “improved customer service, improved employee morale and ongoing improvements.” Suppose customer service did not improve, but did not worsen either. How would a judge or jury determine how much customer service should have improved — and how much DBA had promised it would improve — if DBA had performed all its promises, and what value Rexnord would have derived from such an incremental improvement? The second step would merely be difficult, but the first would be impossible because the provision does not specify the amount of improvement that DBA is obligated to bring about.
The difference between a prediction and a promise is that only the latter is a commitment. Medtech Corp. v. Indiana Ins. Co., 555 N.E.2d 844, 847 (Ind.App.1990); Security Bank & Trust Co. v. Bogard, 494 N.E.2d 965, 969 (Ind.App.1986); Rakestraw v. United Airlines, Inc., 981 F.2d 1524, 1536 (7th Cir.1992); Major Mat Co. v. Monsanto Co., 969 F.2d 579, 583 (7th Cir.1992). A promisor can of course commit to something that is beyond his control and therefore uncertain of attainment. Field Container Corp. v. ICC, 712 F.2d 250, 257 (7th Cir.1983); Oliver Wendell Holmes, Jr., The Common Law 300 (1881). In such a case the commitment operates as a form of unofficial insurance, Cargill, Inc. v. Hardin, 452 F.2d 1154, 1158 (8th Cir.1971), and insurance (spreading risk) is one of the economic functions that contracts perform. E.g., Outboard Marine Corp. v. Babcock Industries, Inc., supra, 106 F.3d at 185; Fidelity & Deposit Co. v. City of Sheboygan Falls, 713 F.2d 1261, 1269 (7th Cir.1983); Jordan v. Group Health Ass’n, 107 F.2d 239, 247-48 (D.C.Cir.1939); State ex rel. Londerholm v. Anderson, 195 Kan. 649, 408 P.2d 864, 875 (1965); see also Spartech Corp. v. Opper, 890 F.2d 949, 955 (7th Cir.1989). But there is no flavor of commitment in the “collateral benefits” paragraph of the consulting agreement. It is merely a prediction that Rexnord will derive unbargained-for as well as bargained-for benefits if it signs the contract. This is further shown by the fact that Rexnord limited its proof of damages, so far as bears on the provision, to evidence of DBA’s having caused Rexnord’s customer service and employee morale to worsen, as a result of DBA’s violating its other undertakings in the contract. That is, Rexnord didn’t attempt the impossible — quantifying the collateral benefits that DBA failed to achieve in its performance of the contract.
*1007True, the instruction relating to the collateral benefits was so broadly worded as to allow the jury to award damages for DBA’s failure to confer those benefits. But DBA does not object to the instruction in this court, and so the point is waived. We add that it is unlikely, though of course not impossible, that the jury awarded damages for which there was no basis in evidence. Juries sometimes do crazy things, but there is no presumption that they do. DBA did object to some of the damages exhibits because they failed (it argued) to establish any causal relation between breach and damages, but it neglected to argue that the failure resulted from the fact that the only “breach” pertinent to those damages was the “breach” of a so-called “promise,” the “promise” of collateral benefits, that was not legally enforceable.
Other issues are raised but do not warrant discussion. The judgment in favor of Rexnord is
Affirmed.