Smith, Kaufman, and Feinberg, Circuit Judges.
The National Labor Relations Act (NLRA), conceived during the Great Depression and founded upon a frank recognition that our boom-and-bust economy was attributable in part to labormanagement unrest, was designed to overcome the inequality of bargaining power between employees and employers. Moreover, the NLRA undertook to provide several methods for the peaceful settlement of labor disputes and the avoidance of disruptions and dislocations generated by strikes and other forms of industrial strife so harmful to our economy. See NLRA section 1, 29 U.S.C. § 151. At the heart of the statutory scheme was the duty, now imposed upon both sides, to bargain with each other "in good faith." NLRA sections 8(a)(5), 8(b)(3) and 8(d), 29 U.S.C. §§ 158(a)(5), 158(b)(3) and 158(d). But, that bargaining duty was not restricted to the negotiation and execution of labor-management agreements; it encompassed post-contractual obligations as well, such as the good faith day-to-day administration of the grievance procedures the parties created for the settlement of their disputes. See United Steelworkers of America v. Warrior and Gulf Nav. Co., 363 U.S. 574, 4 L. Ed. 2d 1409, 80 S. Ct. 1347 (1960). Generally, the National Labor Relations Board (Board) and the Courts have abstained from interfering with the conduct of this dispute-resolving machinery; but, there are occasions, as in the case before us, where we are compelled, in the interest of avoiding the dissension which flows from uncertainty, to set forth guidelines or place our imprimatur on procedures sought to be utilized.
This case is before us on cross-petitions for review and enforcement of an order of the Board directing the Fafnir Bearing Company (Company) to permit the International Union, United Automobile, Aerospace and Agricultural Implement Workers of America, Local 133, UAW, AFL-CIO (Union) to conduct independent time studies at the Company's plant for the purpose of determining whether the Union should accept certain proposed piece rates or proceed to arbitration, the final step in the grievance procedure. The Union, which filed the underlying charge, has intervened in this proceeding.*fn1 For the reasons set forth below, we deny the petition for review and enforce the order of the Board.
The factual setting which gave rise to this proceeding is, in the main, undisputed. Many of the essential facts were stipulated; and, as to other facts, we find substantial evidence to support the findings of the trial examiner and the Board. See NLRB v. Universal Camera Corp., 340 U.S. 474, 95 L. Ed. 456, 71 S. Ct. 456 (1951).
The Company operates a production plant in New Britain, Connecticut which manufactures ball bearings and related products. Since 1944, the Union has represented the Company's production and maintenance employees in periodic negotiations concerning wages, hours and working conditions. According to the pertinent provisions of the contract executed in February 1962, wages are paid pursuant to what is commonly termed a "piece rate system." This method of compensation is designed to provide employees with an incentive to maximize their productivity. Its basic operation is not complex. A time study is conducted by an industrial engineer to determine how many units per hour an average employee working at a normal pace can produce. This figure is called the "standard." Workers who turn out less than or as much as the standard receive a minimum wage. But, the conscientious and resourceful employee, who so desires, can increase his rate of pay by raising the level of his production. If, for example, he produces double the units specified in the applicable standard, he will receive twice as much compensation as his less ambitious colleague.
There is, however, no demonstrably accurate method for setting "standards"; and, as can be readily predicted, this area is a fertile source of disputes requiring invocation of the Company-Union grievance machinery. Indeed, considering that the Company has established over 10,000 new piece work prices each year, the number of grievances to which they gave rise was relatively modest. In 1961, there were 86 piece rate grievances; in 1962, the corresponding figure was 47; and in the first half of 1963, 33 such grievances were filed.
The governing agreement established only general guidelines as to how "standards" were to be formulated. Section 8.4 provided, inter alia, that piece work prices "shall be set so that the average qualified operator working under normal job conditions and applying normal incentive efforts shall be able, after a reasonable trial period, to earn 5% above the standard earning rate of his labor grade (as indicated in the schedule initialled by the parties)." Subsection 8(a)(4) authorized a piece rate revision only when "there has been a measurable change or accumulation of changes over a period of time in material, method or processes not previously taken into account in setting the price . . ."
On various dates prior to February 1963, the Union filed grievances*fn2 claiming that certain piece work prices were improperly established by the Company and did not conform to Section 8.4 of the governing agreement. After unsuccessfully utilizing the first two steps of the Company-Union grievance machinery, key representatives of the Company and the Union met on February 7, 1963, at the third stage of the grievance procedure. At this meeting, the Union requested all time study data which the Company had utilized in setting the piece rates which had prompted the grievances and the Company furnished all the relevant information in its possession. Kermit K. Mead, Director of the Time Study and Industrial Engineering Department of the Union's parent International, the United Automobile Workers, examined this material during a recess, and upon reconvening, proceeded to question Company representatives about the data supplied. Dissatisfied with the responses, particularly because the Company could not account for a 5% "adjustment factor," Mead requested permission to conduct his own time study of the operations involved so that he could verify the Company's proposed rates and intelligently advise the Union whether to accept them or invoke arbitration -- the fourth and final step of the grievance machinery. The Company refused this request on three principal grounds: (1) such study was unnecessary for the Union to determine whether to seek arbitration; (2) the governing agreement did not authorize the Union to conduct independent time studies, and (3) in any event, the Union's rights were adequately protected because the arbitrator would conduct his own studies as he had in the past.*fn3
When the Union filed the underlying charge with the Board, a complaint issued alleging, inter alia, that the Company had refused to bargain in good faith as required by section 8(a)(5) of the Act. Thereafter, a full hearing was held before a trial examiner who found as a fact that Mead was unable to determine "from the data set forth in the Company's studies whether the piece-work rates were established in accordance with the contract, and, therefore, was unable to advise the Union whether to request arbitration." Despite this finding, the examiner believed he was constrained to recommend that the complaint against the Company be dismissed because, inter alia, he was "unable to find any present compelling authority . . . which sanctions the view that a union is entitled to make its own independent time studies after it has received the employer's time study records." The Board*fn4 endorsed its examiner's findings of fact*fn5 but rejected his conclusion. In a determination attacked here by the Company, the Board held that sections 8(a)(1) and 8(a)(5) were violated when the Company refused to grant the Union access to its plant for the purpose of conducting an independent time study of the disputed operations.
A threshold question must be determined by us. Contending that principles of stare decisis should control the outcome of this proceeding, the Company maintains that the decision of the Board is directly contrary to this Court's holding in NLRB v. Otis Elevator Co., 208 F.2d 176 (2d Cir. 1953). While it is true that in Otis, a per curiam opinion (Judge Clark vigorously dissenting), we declined to enforce that portion of the Board's order directing a company to permit a union to conduct an independent time study which the union sought for the purpose of assessing the accuracy of proposed piece work prices, we believe Otis is distinguishable in at least two vital respects. First, the company in Otis not only refused to permit the union to make a "live" study, but, rejected outright the union's request for whatever data the company had in its possession.*fn6 Thus, Otis reached this Court on a sparse record. Neither the trial examiner nor the Board had been afforded an opportunity to analyze the Company's time study material. Consequently, no determination had been made as to whether this data alone would have been sufficient for the union to make an informed judgment to accept the company's proposed rates or seek arbitration. Second, and even more significant, this Court in Otis, on the relatively skimpy record before it, concluded that whatever additional information the union sought could have been obtained merely by interviewing its members. In view of this circumstance, the Court determined that no "invasion of respondent's plant" was necessary. 208 F.2d at 177.
Seizing upon this language, the Company urges that in the instant case, the Union had alternative sources for the information it sought and therefore did not really require independent time studies. But, there was ample evidence before the trial examiner and the Board to justify the rejection of this contention out of hand. Mead testified in detail as to the absolute need for a "live" time study stating that without it the Union could not make an informed judgment on the Company's proposed rates. The testimony of Mead and Bertram Gottlieb, Staff Industrial Engineer of the AFL-CIO, also gave the trial examiner a full view of how time studies were generally made: Piece work prices were established by a ratesetter (usually an industrial engineer) on the basis of a combination of subjective and objective factors. Preliminarily, the ratesetter examined the place where the work was performed and noted environmental factors such as heat, light and other conditions which might have a bearing on productivity. He then reduced the particular job under study into its component parts, designated "cyclic" and "non-cyclic" elements. Cyclic elements were those performed in the production of every unit; non-cyclic elements comprised factors such as going to the lavatory, consulting with the foreman or getting stock from a storage area. Next, the ratesetter observed several operators performing their jobs and made a record of the "actual performance of the job in terms of ...