Nordli, Wilson Associates (“Nordli”) appeals from a decision of the district court refusing to enforce a provision of an employment contract against its former employee, Alan R. Clarke (“Clarke”).
Nordli is a partnership of psychologists serving as professional consultants to businesses and other organizations. Through evaluations (or “assessments”) by tests and interviews, it assists its clients in making judgments about matters such as the hiring, placement, promotion and firing of personnel. There are probably a dozen or two such firms in the country.
Nordli hired Clarke in March, 1964. Then 31 years old, he had received a Ph.D. degree in psychology in 1961; had done post-doctoral work at a state hospital from 1962-3; and had engaged privately in diagnosis and psychotherapy for several educational institutions.
In September, 1964, Clarke signed a letter agreement prepared by Nordli which confirmed his existing employment as a “consulting psychologist” at a starting annual salary of $12,000. The agreement made provision for vacations, sick leave, and fringe benefits. Clarke could leave, or Nordli could terminate him, upon the giving of specified advance notice. If after leaving Clarke went to work for a Nordli client or prospective client, certain obligations were imposed.1 These obligations are the subject of this litigation.
*1220Clarke’s principal duties with Nordli were to make psychological assessments of candidates for positions within its business clientele. Two of these were life insurance companies, for which he mainly assessed persons seeking positions as salesmen (he also assessed some prospects for advancement or termination). For a Buffalo, New York, firm, he assessed candidates for middle or lower-middle management positions, and a few candidates for executive positions. A single assessment would take from one to one-and-a-half hours. Clarke would meet the individual and administer a psychological test designed by Nor-dli. He would later prepare a report “pulling together” (as he testified) the test results and interview impressions, which he would usually submit to the client company’s director of industrial relations or personnel. He might also, in some cases, meet again with the person tested and provide “feedback”, i. e., explain the evaluation to him.
At the time Clarke began working for Nordli, ITT was not its client. In April 1964, Dr. Beare of ITT called Mr. Wilson, a Nordli partner, to arrange an appointment to discuss Nordli’s becoming one of ITT’s consulting firms. Wilson could not keep the appointment, and Clarke met with Beare instead. Thereafter, Nordli worked for ITT “in a small way.” In September, 1965, Dr. Beare called Clarke and asked if he was interested in working for ITT. Beare said that he was planning to leave and was looking for someone to replace him. In October, 1965, Clarke, still in Nordli’s employ, made two assessments of candidates for ITT employment in New York. In early November, 1965, he indicated that he would leave Nordli, and did so early in 1966.
Nordli contends that Clarke’s duties with ITT (until August 1969, when it concedes they became “very much different”) were such that the latter received “professional psychological services” from Clarke within the terms of the employment contract. As ITT had been its client, Nordli claims that it is entitled to 15% of Clarke’s ITT compensation from salary and bonuses from February 1, 1966 until August 1969. Clarke denies any such liability.
Clarke’s title at ITT was “Manager, Executive Selection and Appraisal.” In September, 1966, it was changed to “Director of the Executive Selection and Appraisal Branch.” Clarke administered and supervised ITT’s “assessment center” in New York. His staff consisted of one (later two) psychologists, an administrator, and clerical personnel. The center, headed by Clarke, hired outside psychological consultants, like Nor-dli, which under its supervision carried out assessments all over the world of candidates for ITT executive positions. The aggregate fees paid to outside consultants were ten times the budget for the internal employees supervised by Clarke. The assessment center’s function appears to have been to establish standard tests and procedures, coordinate testing, and arrange and supervise a unified program carried out largely by the hired consultants.
From Clarke’s testimony, which the district court found to be truthful, it could reasonably be found that the preparation of test materials and the coordination of the activities and programs of the assessment center was carried out by others than Clarke, but under his overall direction. Written tests were prepared by the man responsible for administration of the center. They were approved by Clarke but not altered by him. The tests were standard psychological tests. They were not the tests used by Nordli (Clarke testified in his deposition to dissatisfaction with the Nordli tests). Interview guidelines were “typically” given by the administrator of the center. Communications with consulting psy*1221chologists were normally through the staff psychologist, who from time to time consulted with Clarke. Clarke did not usually read assessment reports. Clarke denied developing any new assessment concepts and practices and denied maintaining quality control of reports, although he admitted that these matters were within his overall responsibility. Statistical studies of the effectiveness of assessments were maintained by a subordinate. There was no evidence that anything done by Clarke at ITT was copied from practices uniquely developed by Nordli.
In 1967, when Clark’s title changed, he delegated more of the responsibilities to subordinates. Much of his time was spent developing a “rather ambitious” program of executive training.
Clarke testified that he himself conducted a few individual assessments and feedbacks prior to 1967 when subordinates or consultants were not available, but these consumed less than one per cent of his time. He maintained that these were not psychological services, since assessments were often performed by non-psychologists. He was not a licensed psychologist in New York state, although a license would be required in order to practice there. He stated that he had left Nordli because of dissatisfaction with the role of a psychologist, his interests lying more in administration and business organizations.
The district court concluded that from February 1, 1966 to June 1967, Clarke spent less than one per cent of his time with ITT on professional psychological work, and that since June 1, 1967, none of his ITT activities included or constituted professional psychological work, being more properly characterized as personnel and administrative work.
In August of 1966, Nordli notified ITT that it would no longer provide it with psychological consulting services (which it said it had provided for “about a year”). In its letter, Nordli alluded to alleged unethical practices involving the pirating of employees, including Clarke. The district court found that Nordli had exacerbated its damages, if any, by “firing” ITT.
Nordli challenges the district court’s conclusion that only Clarke’s very occasional (less than 1%) personal participation in assessments and feedbacks constituted professional psychological services to ITT within the meaning of the contract. It asserts that Clarke’s responsibility for the global activities of outside psychologists hired and coordinated by the assessment center, which he headed, constituted the furnishing “directly or indirectly” of “professional psychological services of any sort” to ITT.
We hold, however, that even were the agreement to be so interpreted, Nordli may not recover. It would be unenforceable both as a restraint going beyond Nordli’s legitimate interest in its own self-protection, and as imposing liquidated damages insufficiently related to any harm to Nordli.
Bargains between employers and employees which place restrictions upon an employee’s exercise of a gainful occupation after leaving the employer are, and were at common law considered to be, void as in restraint of trade. Restatement of Contracts, §§ 513-515. Sherman v. Pfefferkorn, et al., 241 Mass. 468, 474, 135 N.E. 568, 570 (1922). Exceptions are made, however, if the restriction is reasonable, “and not wider than is necessary for the protection to which the employer is entitled. .” Sherman v. Pfefferkorn, supra at 474, 135 N.E. at 569.
In cases of agreements forbidding an employee to engage in the same business or line of activity as the employer, attention. is often given to whether the agreement is reasonably limited as to area and duration. But such considerations are merely aspects of the rule, quoted above, that the agreement must be no wider than is necessary to afford reasonable protection to the employer. A further, and here more relevant, aspect of the rule is that an employer may *1222not enforce limitations on an ex-employee’s engaging in activities which do no damage to the employer. Novelty Bias Binding Co. v. Shevrin, 342 Mass. 714, 175 N.E.2d 374 (1961), and cases cited therein. See also Baker v. Starkey, 259 Iowa 480, 144 N.W.2d 889 (1966) involving an employee who left a management consultant group.
In two recent decisions, the Massachusetts Court has refused to enforce non-competition covenants in employment agreements where the plaintiff-employer could not prove that it was damaged by the employee’s work for another firm. Richmond Brothers, Inc. v. Westinghouse Broadcasting Company, Inc., 357 Mass. 106, 256 N.E.2d 304 (1970); Slade Gorton & Co. v. O’Neil, 355 Mass. 4, 242 N.E.2d 551 (1968).
Clarke’s duties at ITT, whether or not designated “professional psychological services”, did not injure Nordli by diminishing or dispensing with the need for Nordli’s services. Assuming, as we do, that Nordli had every right to protect itself reasonably against Clarke’s utilizing his position at Nordli to steal a Nordli client, the facts here do not indicate that that is what happened. It is true that Clarke’s Nordli position was a springboard from which his ITT association was launched; and had he left Nordli to become an independent (or ITT hired) consulting psychologist doing assessments for ITT in competition with Nordli, the situation would be different. However, Clarke took over a management function at ITT. The work he managed included the making of assessments. But his role was not that of a consulting psychologist. He was the liaison between higher ITT management and the assessment of ITT executives. He directed and hired people to do what he had once done.
There is no evidence that the Nordli firm itself would have been a candidate for Clarke’s executive position, or that Clarke’s very occasional fill-in activities in making assessments in any way diminished ITT’s continued reliance upon independent psychological consultants. The vast bulk of ITT’s assessments were performed by independent firms, like Nordli. Indeed, Nordli remained an ITT consultant until it initiated its own withdrawal.
We recognize that Nordli’s express reason for withdrawal was what it felt to be unethical practice by ITT (and, perhaps, friction with Clarke resulting from his apparently adverse appraisal of Nordli’s testing procedures); but these matters do not demonstrate that Clarke’s ITT role placed him in competition with Nordli. (Indeed, since ITT’s assessments were performed by outsiders, Clarke’s new role was potentially beneficial to his former employer had they remained on good terms.) The point is that Clarke did not, by reason of his new duties, displace or compete with Nordli, nor is there evidence that he damaged Nordli by using confidential information (such as its tests) for the benefit of others. Like the Supreme Judicial Court in Richmond Brothers, Inc. v. Westinghouse Broadcasting Company, Inc., supra, we are of the opinion that enforcement of the terms of the agreement is not “reasonably necessary for the protection of the plaintiff’s business.”
Perhaps because it recognizes the absence of damage based on unfair competition, Nordli stresses its damages allegedly suffered by losing a valuable employee whom it had trained. (This damage would, of course, be the same had Clarke left to go to work for a non-client instead of for a client or prospective client. It would likewise be the same whether he found employment as a psychologist or in some other capacity.)
An employer’s right, however, to place future burdens on an employee merely to forestall or offset the loss of his services is clearly limited:
[A]n employer cannot by contract prevent his employee from using the skill and intelligence acquired or increased and improved through experience or through instruction received in the course of the employment. The *1223employee may achieve superiority in his particular department by every lawful means at hand, and then, upon the rightful termination of his contract for service, use that superiority for the benefit of rivals in trade of his former employer. As was said in Herbert Morris, Limited, v. Saxelby, [1916] 1 A.C. 688, at 714: ‘A man’s aptitudes, his skill, his dexterity, his manual or mental ability ought not to be relinquished by the servant; they are not his master’s property; they are his own property; they are himself.’
Club Aluminum Co. v. Young, 263 Mass. 223, 226-227, 160 N.E. 804, 806. See also Richmond Brothers, Inc. v. Westinghouse Broadcasting Co., Inc., supra, 357 Mass, at 111, 256 N.E.2d 304.
Doubtless an employer who has provided specialized training to an employee — as by a course of studies or the like — might reasonably contract with the employee for reimbursement if the employee should quit before the employer achieves any benefit. However, the employer may not require its ex-employee to make payments to it unrelated to the employer’s damage, simply as a penalty to discourage or punish a job change. At most, the employer, in appropriate circumstances, might require the employee to reimburse it for a sum, or under a formula, reasonably related to what it cost the employer to train him, or to retrain a replacement, or the like. The.sum or formula would be enforceable only if “the amount so fixed is a reasonable forecast of just compensation for the harm that is caused” by the loss of the employee — to borrow words from Section 339 of the Restatement of Contracts, setting forth the requirements for an enforceable liquidated damages provision.
Viewed as an advance effort to liquidate Nordli’s actual damages caused by the loss of an employee, the 15% clause is not impressive. We perceive no rational connection between 15% of Clarke’s ITT compensation (which Nor-dli tells us is $24,525 for the five years from 1966-1970, and $14,962.50 for the shorter period now claimed) and whatever damages Nordli sustained by reason of Clarke’s departure after about two years with Nordli.
We recognize that a liquidated damage provision is appropriate where the harm is incapable or very difficult of accurate estimation. Restatement of Contracts, § 339(1)(b). But the plaintiff had the obligation of demonstrating some basis upon which, when the formula was fixed, it could be said to be a reasonable forecast of damages. We can find none.2 There is no evidence of what it cost to train someone like Clarke, or of Nordli’s costs occasioned by the loss of such an employee. There was no evidence of what training, if any, was needed or was received by Clarke in order to evaluate tests and conduct assessments, over and above his previous education and experience.
Moreover, the formula itself is plainly deficient viewed as an attempted forecast of Nordli’s damages from the loss of an employee. By its terms, it applies whether Clarke had worked for Nordli for ten years or two years, even though in the former case their investment in him would presumably have been repaid. It makes no distinction between damages occasioned by mere loss of services and the additional damages that Nordli would have been caused had Clarke engaged in direct competition. It applies whether Clarke voluntarily quit or was fired.3
*1224Accordingly, we decline to enforce the contract against Clarke. If read as appellant contends, we find it to be an unreasonable restraint; and, of course, if not so read, it would be inapplicable.4
Affirmed.