Opinion for the Court filed by
LEVEN-THAL, Circuit Judge.Like UAW v. Marshall, 189 U.S.App.D.C. 232, 584 F.2d 390 (1978), also decided today, this case involves a challenge to a decision of the Secretary of Labor denying a request for certification for worker adjustment assistance under the Trade Act of 1974. At issue is interpretation of the phrase “contributed importantly” as used in § 222(3) of the Trade Act. That subsection states:
The Secretary shall certify a group of workers as eligible to apply for adjustment assistance under this part if he determines—
(3) that increases of imports of articles like or directly competitive with articles produced by such workers’ firm or an appropriate subdivision thereof contributed importantly to such total or partial separation, or threat thereof, and to such decline in sales or production.1
We affirm the decision of the Secretary denying certification.
I. BACKGROUND AND CONTENTIONS ON APPEAL
On June 13, 1976, the Mount Vernon, Ohio, sheet glass manufacturing facilities of Pittsburgh Plate Glass (PPG) Industries, Inc., were closed and the workers laid off for an indefinite period. On June 21, 1976, the Stone, Glass and Clay Coordinating Committee, on behalf of the unions representing the Mount Vernon workers, submitted to the Secretary of Labor a petition for eligibility to apply for worker adjustment assistance. The Certifying Officer, acting under delegation by the Secretary, denied the petition, concluding that increased imports did not contribute importantly to separations and the decline in sales and production at the Mount Vernon facility.2 The Committee petitioned for reconsideration of that determination. After a second investigation, the Secretary again declined to certify the workers at the Mount Vernon plant. The unions filed a petition in this court to review that order.3
The petitioners claim that in concluding that imports did not contribute importantly to separations at the Mount Vernon plant, the Secretary misinterpreted the phrase “contributed importantly” and made findings not based on substantial evidence. They maintain, indeed, that properly *242viewed through the lens of the statute, the Secretary’s own findings establish that increased imports did contribute importantly to the Mount Vernon separations. They further contend that the investigative methodology employed by the Office of Trade Adjustment Assistance resulted in an understating of the impact of imports on the Mount Vernon facility. Our analysis of the statute and the Secretary’s approach leads us to reject these contentions.
A. The Meaning of “Contributed Importantly” as used in the Trade Act of 1974
As this court notes in UAW v. Marshall, supra, the legislative history of the Trade Act of 1974 indicates that Congress chose the phrase “contributed importantly” in a conscious effort to make adjustment assistance more readily available than it had been under the Trade Expansion Act of 1962. In the earlier act the requirement had been that imports be shown to be a “major cause” of the separations.4 “ ‘Major’ ha[d] been understood to mean greater than all other factors combined.”5 In contrast, § 222 of the 1974 Trade Act defines a cause that “contributes importantly” as “a cause which is important but not necessarily more important than any other cause.”
The legislative history gives very little guidance on the meaning of “importantly” in this context. The pertinent Senate committee report observed:
A cause must be significantly more than de minimis to have contributed importantly, but the Committee does not believe that any mechanical designation such as a percentage of causation can be realistically applied.6
That report went on to observe that “total or partial separations that would have occurred regardless of the level of imports, e. g., those resulting from domestic competition, seasonal, cyclical, or technological factors are not intended to be covered by the program.”7
Congress therefore did not afford a precise definition of the phrase “contributed importantly.” Petitioners attempt to divine more precision by invoking a number of analogies, all aimed at associating a percentage of causality with the term “important.” For example, petitioners note that one criterion for worker adjustment assistance certification is “that a significant number or-proportion of the workers . become totally or partially separated.”8 One meaning of “significant” is “important.” Petitioners argue that Congress intended the term “significant” to encompass as little as “5 percent of the workers” at a firm.9 They contend that a cause of a similar relative magnitude must be deemed to be one that “contributed importantly.”
Petitioners also point to findings of the International Trade Commission (ITC), formerly the United States Trade Commission. Under the Trade Act of 1974, the ITC is responsible for investigating “whether an article is being imported into the United States in such increased quantities as to be a substantial cause of serious injury, or the threat thereof, to the domestic industry producing an article like or directly compet*243itive with the imported article.”10 The Commission is to report its findings to the President, who is to provide import relief in the form of duties, tariff-rate quotas, quantitative restrictions, or orderly market agreements if he finds such relief is in the public interest.11 Petitioners focus on the ITC’s interpretation of “a substantia] cause of serious injury” because that encompasses a finding that the cause is “important,” which in turn bears some similarity to a determination that a cause “contributes importantly.” 12 Petitioners conclude:
On several occasions, the Commission has determined that increased imports were an important cause of injury, though not necessarily a coequal cause, even though imports were only one of several causes and, statistically, imports increased their market penetration by less than 10 percent and captured less than 20 percent of the domestic market. Besides both indirect and direct evidence of lost sales, the Commission also looks to the price-suppressing influence exerted by lower priced imports.13
Finally, petitioners contend that Congress intended relief to be available, even where the impact of imports is ambiguous. In support of this proposition, they note that imports need only increase relatively to meet the “increased imports” requirement — clearly Congress’s attempt to facilitate a finding of causality notwithstanding a plethora of reasons for separations in a declining market.14
These lines of analysis have some bearing, but they cannot prevail if we accept the Secretary’s key finding — that the relative importance of imports is fairly captured in the fact that customers who decreased purchases from Mount Vernon and increased purchases of imports represented less than 8% and 4% of surveyed Mount Vernon sales in 1974 and 1975, respectively.
Petitioners apparently concede that if this finding stands, the court cannot rule as a matter of law that the “contributed importantly” criterion is satisfied.15 Therefore, as a crucial coordinate contention, petitioners maintain that the Secretary’s factual findings reflect a lack of understanding of the statutory criteria and mis*244characterize the data. We turn to the Secretary’s factual findings and investigative methodology.
B. Investigative Methodology
1. The OTAA Studies
The Office of Trade Adjustment Assistance (OTAA) of the Department of Labor is responsible for investigating worker adjustment assistance petitions.16 It conducted two investigations regarding the workers at the Mount Vernon plant. In the first investigation, OTAA included a trade and industry analysis that identified a number of salient characteristics of, and trends in, the sheet glass industry and the larger, encompassing flat glass industry. These characteristics and trends may be summarized as follows:
1) The flat glass industry is composed of producers of sheet glass, float glass and plate glass. These glass types are products of different manufacturing processes, but are highly substitutable in application.
2) Recent technological progress in the manufacture of float glass has led to substantial displacement of plate and heavy sheet glass by float glass in the domestic market.17 The sheet glass share of the domestic production of flat glass declined from 24.4% in 1974 to 16.3% in 1975.
3) The drop in domestic automobile production and in building construction during the 1973-1975 period had serious adverse effects on the domestic glass industry.18
OTAA then went on to look at specific circumstances at the Mount Vernon plant. It conducted a survey of customers of the Mount Vernon plant as well as domestic users of sheet glass who purchased their requirements elsewhere. Twelve customers of the Mount Vernon plant, representing 36.2% of its sales, were questioned to determine whether they were decreasing their purchases from PPG and increasing their purchases of imports. Eight of the 12 customers stated that they had increased their purchases from the Mount Vernon plant; three stated that they had shifted to float glass. One customer responded that it switched to other domestic manufacturers. Seven of the 12 surveyed customers indicated that they either did not import, imported very little, or had been decreasing their level of imports. Four customers did not state whether they imported. One customer indicated that it increased purchases both from PPG and of imports.
OTAA contacted 13 residential and storm window manufacturers to determine the extent to which they import. At the time of the survey, none of these manufacturers imported the type of glass produced at the Mount Vernon plant. Four indicated that they had imported in the past, three had stopped three years ago and one stopped the previous year. The survey did reveal that all manufacturers were purchasing substantial quantities of float glass — up to 50% or more of their glass requirements.
On the basis of the collected information, OTAA recommended that the petition for certification of the workers at the Mount Vernon plant be denied.19 The affected *245unions then filed a petition for reconsideration.
After the petition for reconsideration was granted, OTAA conducted a second study.20 It focused solely on the impact of imports on the Mount Vernon plant. The study included a survey of 45 domestic users of sheet glass in order to discern their purchasing patterns. Of the firms surveyed, 21 were customers of the Mount Vernon plant and accounted for 21.5% of the plant’s sales in 1974 and 38.3% in 1975. Of these, customers accounting for 13.5% of Mount Vernon’s sales in 1974 and 13.8% in 1975 decreased their purchases from Mount Vernon over the 1974 to 1975 period and increased (absolutely and/or relatively) their purchases from other domestic suppliers. A few customers decreased purchases from Mount Vernon and did not purchase sheet glass from any other source. Customers accounting for 6% of sales in 1974 and 21.3% in 1975 increased their purchases from the Mount Vernon plant absolutely or relatively from 1974 to 1975.
Most significantly, the second investigation revealed that customers who decreased purchases from Mount Vernon and increased purchases of imported sheet glass over the 1974-1975 period represented only 1.6% of Mount Vernon sales in 1974 and 1.5% of Mount Vernon sales in 1975.
Since Mount Vernon customers representing over 90% of the sales reflected in the sample “either switched to other domestic competitors, reduced purchases from all sources or increased purchases from the Mt. Vernon plant,” while those representing less than 8% “decreased purchases from PPG and increased purchases of imports of sheet glass,” the OTAA recommended denial and the Secretary consequently concluded that increases in imports had not contributed importantly to separations at the Mount Vernon plant.21 The Memorandum of Recommendation noted that the sample of 21 customers included names submitted to OTAA by PPG officials and petitioners in the belief that those customers were importing sheet glass. Thus, the Memorandum concluded, the sample may be biased in favor of customers who import sheet glass.22
2. Petitioners’ Objections
The petitioners maintain that the implicit “important cause” test underlying OTAA’s studies and the Secretary’s conclusions is that customers of the affected firm must (1) decrease purchases from the affected firm and (2) increase purchases of imported articles, before imports may be found to have contributed importantly to separations. They argue that this dual test is not in the statute and was not intended by Congress; that it obscures the true effect of import competition and totally ignores the effects of price suppression.
The petitioners’ core complaint is that simply because old customers reduced purchases from Mount Vernon and increased purchases from other domestic suppliers does not mean that the lost sales were not causally related to an increase in imports. Their argument appears to be that the effect of intense import competition is felt throughout the industry. Domestic firms cut prices to maintain their market shares. The end result is that a marginal facility, such as the Mount Vernon plant, loses customers to foreign competition both directly and indirectly — indirectly in that other domestic firms who have lost customers to imports will compete more intensely for the customers of other domestic suppliers. The *246net result is that the marginal firm in the industry — i. e., the firm with the highest production costs — is the big loser.
Closely related to this argument are petitioners’ contentions regarding price suppression. Price suppression occurs when low-priced imports force domestic producers to cut prices to artificially low levels. Again, the marginal firm is hit the hardest. The petitioners claim that the effects of price suppression were given no meaningful attention in the OTAA studies.
The petitioners’ final differences with OTAA’s studies are with the actual mechanics of counting up the percentages relied on by the Secretary. To use the unions’ language, they claim “the Secretary couldn’t count straight.”23 As noted above, the ultimate finding of OTAA’s second investigation was that Mount Vernon customers representing over 90% of the sales in the sample “either switched to other domestic competitors, reduced purchases from all sources or increased purchases from the Mt. Vernon plant,” and less than 8% “decreased purchases from PPG and increased purchases of imports of sheet glass.” The unions contest the computational method by which these numbers were arrived at.24 The basic methodology consists of identifying those customers who increased foreign purchases in a given year, and calculating their Mount Vernon purchases as a percentage of total Mount Vernon purchases by all customers included in the survey. This calculation was used by OTAA as an indicator whether substantial Mount Vernon customers were increasing purchases of imports while decreasing purchases from Mount Vernon.
' The technique is crude and petitioners are quick to point to some of the anomalies presented by this method of computation. For example, one former customer of Mount Vernon who was part of the survey purchased nothing from that plant in 1975, and so was not reflected in the 1975 computation, yet did purchase more imports in that year than in 1974. The petitioners also note that this method of computation does not reflect sales lost to imports where the customer simply maintained the normal level of Mount Vernon purchases while increasing his overall purchases of sheet glass, satisfying the new increment with imports. Petitioners conclude that a method of computation that does not reflect such obvious shifts to imports is unreasonable.
II. THE VALIDITY OF THE DENIAL OF ADJUSTMENT ASSISTANCE
In UAW v. Marshall, this court has noted the general remedial purposes of the Trade Act of 1974 and how those purposes affect the role of the reviewing court. In that case we were concerned with an interpretation of the statutory phrase “an appropriate subdivision”, which had never been explained by the Secretary and which had substantial ramifications not only in the case at issue, but for all petitions for worker adjustment assistance. In light of the remedial purposes of the Trade Act, we found that the failure to articulate the reasoning behind the Secretary’s rigid interpretation was unreasonable.
This case touches on the Secretary’s interpretation of the Act, but is more directly an attack on the investigative methodology employed by OTAA. OTAA is required to investigate a wide range of industries and to some extent the technique by which it attempts to answer the question whether an increase in imports “contributed importantly” to separations will vary with the structure of the industry, the available sources of information, and the number of other causative factors at work. Methodological flexibility is critical if the Secretary *247is to discharge his official responsibility under the worker adjustment provisions of the Trade Act. Consequently, this court must show substantial deference to the agency’s chosen technique, only remanding a case if that technique is so marred that the Secretary’s finding is arbitrary or of such a nature that it could not be based on “substantial evidence.”25
We first consider whether the methodology in this ease was so defective that the Secretary’s conclusion, regarding the relative weight of imports among the operative causes of separations, is not supported by substantial evidence. We then address the core statutory issue: given the relative importance of imports as determined by OTAA, was the Secretary’s finding that imports did not “contribute importantly” to separations consistent with the statute?
A. Methodology
OTAA’s first report painted a portrait of a small, marginal firm swamped by technological obsolescence and an economic recession. Advances in the production of float glass have made it more attractive than sheet glass in many applications. The inroads made by float glass and the impact of the recession in the automobile and construction industries are not seriously challenged by petitioners. Their claim is that OTAA’s methodology seriously understates the role of imports as another important contributing factor.
The complaint thus focuses on what petitioners characterize as the Secretary’s “dual test”: Did customers reduce purchases from Mount Vernon and increase purchases of imports? The Secretary maintains that such a test was necessary to sort out any causal link between imports and separations. We think that in the circumstances of this case the Secretary’s position is a reasonable one.
The Secretary was confronted with a situation in which there were two dominant agents in Mount Vernon’s decline: a deep recession and the accelerating technological obsolescence of the Mount Vernon plant, partly reflected in float glass’s superiority in terms of price and performance.26 He had to filter out the contribution of imports, which had actually been declining in volume in recent years.
Admittedly, looking to whether a customer had directly shifted to imports is not a very sophisticated test. Imports can have a more subtle impact on purchasing patterns. If cheap imports have forced a domestic rival to cut prices to a level the marginal domestic firm cannot sustain, imports have definitely hurt the marginal firm although on paper some or all of its customers might have shifted to the domestic rival. But in the instant case such indirect effects would be extremely hard to trace because of the general decline in demand due to the reces*248sion and competition from float glass.27 Moreover, according to the survey responses, a number of customers who switched to other domestic suppliers did so for technological reasons apparently unrelated to price: e. g., a switch to a different glass thickness, or to plastic, polyethylene, or tempered safety glass.
Even if this court’s review function permitted it, we would be hard pressed to suggest a better test. Nor have petitioners done so. A court must take into account that in many instances of economic analysis, it is possible to identify flaws in the methodology used by an agency without being able to suggest a substitute that is better.28
The Secretary was also well within his discretion in excluding as evidence of the impact of imports those customers who are increasing purchases of imports while maintaining their purchases from Mount Vernon. It may be assumed, as petitioners argue, that if those customers had not fulfilled their growing demand with imports, these additional purchases would have gone to domestic firms. However, the provision of the Trade Act that relates to compensable impact focuses on the loss of established customers and on the failure to maintain past sales levels.29 In sum, we find the “dual test” applied by the Secretary a reasonable means of ascertaining a causal link between imports and separations at the Mount Vernon plant.
Finally, we are left with petitioners’ contention that even if one accepts the dual test, the Secretary and OTAA did not apply it properly. They specifically refer to one customer who is part of the 1974 data, but who does not appear in the 1975 data as a customer increasing imports even though it increased its purchase of imports 200% in that year. The reason for the omission is the customer’s failure to make any purchases from Mount Vernon in 1975.
This flaw in the method by wffich OTAA expressed the impact of imports is troublesome. On the other hand, the issue is not rigorously pressed, being used by petitioners more for color than as a fundamental challenge, and it is not squarely joined by the government. In our own reflection on the problem, we discern that the scheme of the statute may not invite a rigorous historical approach, unearthing all “former” customers who have switched to imports and thus contributed in some way to separations. For § 223 of the Trade Act limits assistance to workers separated no more than one year before the date of the petition for certification.30 This has led the Secretary to confine himself to two years of import data in some worker adjustment assistance cases. The practice was recently discussed by the Seventh Circuit in Paden v. United States Department of Labor, 562 F.2d 470, 473-74 (7th Cir. 1977):
The Secretary submits that by confining consideration to imports during the year of separation and the immediately preceding year, the Secretary can focus on those imports which are most likely to affect employment in the year of separa*249tion while diminishing consideration of those factors which, while affecting employment, are not within coverage of the act. Although petitioners’ contention that imports in any given year may adversely affect employment several years later is not without validity and is indeed recognized as a distinct possibility by the Secretary, we are not convinced that the Secretary’s practice should be condemned either generally or under the circumstances of this particular case.
The Act does not embrace a subdued historical approach, one that takes into account those who were customers but have switched to imports in the two years before the certification petition was filed. But these apparently do not loom large in the picture so far as petitioners are concerned. OTAA computed both 1974 and 1975 percentages. The one 1974 customer who did not purchase from Mount Vernon in 1975 accounted for only a small percentage of Mount Vernon’s sales in 1974, .04%, and only .18% of the 1974 sample of sales.
In sum, we conclude that OTAA’s investigations yielded a picture of the relative importance of imports that is reasonably, though not 100%, accurate. Customers who decreased purchases from Mould Vernon and increased (absolutely or relatively) purchases of imports represented only 7.4% of 1974 Mount Vernon sales included in the survey and only 3.9% of 1975 surveyed sales — despite the fact that the survey was supplemented with customers who petitioners believed would substantiate their claims.
B. The Statutory Standard
The final question then is whether, as a matter of statutory interpretation, the Secretary erred in concluding from this evidence that imports had not contributed importantly to separations. The most salient portions of the legislative history are those that eschew any “mechanical designation such as a percentage of causation,”31 and which admonish that adjustment assistance is not to be made available for separations that result “from domestic competition, seasonal, cyclical, or technological factors.”32
Congress was concerned that the Trade Act’s worker adjustment assistance provisions not become a general program of unemployment assistance. The Secretary was required to find an important causal nexus between imports and separations. The meaning of “important” is not susceptible to any simple definition and Congress did not try to articulate one. As with many governmental programs, Congress looked to the administrator to develop some “feel” for the standard, and consistency in application, through the actual process of determining cases. In such circumstances the reviewing court is chiefly a bulwark against arbitrariness, whether it manifest itself in deviations from an ascertainable legislative mandate, unexplained rigidity in the exercise of the agency’s interpretative function,33 or inconsistency in actual application without reasoned explanation. But a court must afford the agency substantial deference in dealing with complex and diverse applications under an admittedly vague mandate.
What we find here is an ill-defined legislative mandate, and we cannot say that the Secretary has defied it or deviated from its ascertainable meaning. Congress has noted the futility of attempts to define “important” in terms of a “percentage of causality.” There is evidence in the legislative history that Congress believed that even where imports were at least as important as any other cause, the Secretary might reasonably conclude that imports were not an “important” cause.34 The Secretary was within his discretion in concluding that the move to imports found among customers representing 4% to 8% of the sampled sales — which we have sustained as adequately capturing the relative signifi*250canee of imports — was not “important” within the meaning of the Act.
Petitioners offer the thought that they should prevail where the impact of imports is “ambiguous.”35 That borders on the assertion of a presumption, somewhat delicately phrased, which would have to be reconciled with the affirmative duty that the 1974 Trade Act places on the Secretary to find a significant causal relationship. But we need not decide whether, when, or to what extent such a presumption may be invoked. In this case the Secretary concluded that the impact of imports was reasonably measurable, rather than ambiguous, and further concluded that it was too light in the scale.
Affirmed.
. 19 U.S.C. § 2272(3) (1976) (emphasis added).
. 41 Fed.Reg. 41985 (1976).
.The petitioners are the United Glass and Ceramic Workers of North America, AFL-CIO, the Glass Bottle Blowers Association of the United States and Canada, AFL-CIO, and the Stone, Glass and Clay Coordinating Committee.
. Trade Expansion Act of 1962, Pub.L.No. 87-794, § 301(c)(3), 76 Stat. 884.
. H.R.Rep.No.93-571, 93d Cong., 1st Sess. 46 (1973).
. S.Rep.No.93-1298, 93d Cong., 2d Sess. 133 (1974), reprinted in [1974] U.S.Code Cong. & Admin.News, pp. 7186, 7275.
. Id., reprinted in [1974] U.S.Code Cong. & Admin.News at 7275. See also The Trade Reform Act of 1973: Hearings on H.R. 10710 Before the Senate Comm, on Finance, 93d Cong., 2d Sess., pt. 2, at 396 (1974) (Statement of Peter J. Brennan, Secretary of Labor):
Under the revised procedures of H.R. 10710 [the bill that became the Trade Act of 1974], it is not our intention to provide trade adjustment assistance to workers whose unemployment, or underemployment, is clearly the result of normal seasonal or cyclical factors, or of shifts in technology or of domestic competition. Our regular unemployment insurance and manpower programs are designed to deal with such displacement problems.
. 19 U.S.C. § 2272(1) (1976) (emphasis added).
. S.Rep.No.93-1298, 93d Cong., 2d Sess. 133, reprinted in [1974] U.S.Code Cong. & Admin. News, pp. 7186, 7275; H.R.Rep.No.93-571, 93d Cong., 1st Sess. 54 (1973).
. 19 U.S.C. § 2251(b)(1) (1976) (emphasis added).
. Id. §§ 2252(a)(1)(A), 2253(a).
. A “substantial cause” is defined in § 201(b)(4), 19 U.S.C. § 2251(b)(4) (1976), as “a cause which is important and not less than any other cause.” In explaining the difference between the “substantial cause” and “contributed importantly” criteria, the House Ways and Means Committee made the following observation:
The requirement that import increases contribute “importantly” may be contrasted with the “substantial cause” language in the import relief section of the bill. “Substantial cause” in determining eligibility for import relief includes the concept “important” but adds another requirement, that the cause be not less than any other single cause. Therefore, importantly as used in determining eligibility for worker adjustment assistance is an easier standard; a cause may have contributed importantly even though it contributed less than another single cause.
H.R.Rep.No.93-571, 93d Cong., 1st Sess. 53-54 (1973).
. Brief for Petitioners at 8.
. The House version of § 222 of the Trade Act had required an actual or relative increase in imports. The Senate Finance Committee was apparently unhappy with the laxity of the relative increase standard. In a declining market, the overall decline in demand for a certain product is often the predominant cause of separations. So the Senate Finance Committee reported out a version of the bill that required a showing of an absolute increase in imports.
To this end it is required that imports increase absolutely — since it is more likely to be the case under conditions of absolute increases of imports of like and directly competitive products that imports would contribute importantly to the total or partial separation of workers or the threat of such total or partial separation.
S.Rep.No.93-1298, 93d Cong., 2d Sess. 133 (1974), reprinted in [1974] U.S.Code Cong. & Admin.News, pp. 7186, 7275.
In conference, the Senate receded from its position. H.R.Rep.No.93-1644 (Conf.Rep.), 93d Cong., 2d Sess. 35-36 (1974), reprinted in [1974] U.S.Code Cong. & Admin.News, pp. 7367, 7381.
.Brief for Petitioners at 45: “Undoubtedly, figures as low as 7.4 and 3.9 percent would appear to be outside Petitioners’ asserted interpretation of ‘contributed importantly.’ ”
. 29 C.F.R. § 90.12 (1977).
. According to the first OTAA report, Worker Investigation TA-W-943, float glass is now superior in quality and less expensive to produce. Appendix at 10.
. There were declines in both public and private construction during this period. The decline in public construction was the result of the recession and financial pressures on federal, state, and municipal budgets. The drop in housing construction was due to a very tight mortgage funds market, increased costs of housing construction and maintenance, a drop in real income, rising property taxes, and general uncertainty generated by the energy crisis and the economic downturn. Appendix at 14.
.OTAA made the following findings:
1. The recent technological advances making possible the production of lighter weight window glass by the float process has enabled glass manufacturing firms to substitute the better quality float glass into markets formerly dominated by sheet glass.
2. Customers of PPG Industries, Inc. who have decreased purchases of sheet glass manufactured at the Mt. Vernon plant have switched to purchasing domestically manu*245factured float glass and have not purchased imported glass.
3. The new home construction industry is a primary market for the type of glass manufactured at the Mt. Vernon plant of PPG Industries, Inc. The depressed condition of the construction industry in the late 1974-1975 period required a reduced supply of sheet glass.
Appendix at 58.
. Id. at 70-155. A summary of the findings of this study can be found in the Appendix at 144.
. Id. at 159-60 (letter from Joel Segall, Deputy Under Secretary, Bureau of International Labor Affairs, to counsel for petitioners).
. Id. at 157.
. Brief for Petitioners at 10.
. The unions also suggest that the more appropriate percentage would be the percentage of customers who decreased purchases from Mount Vernon and increased purchase of imports, without weighting the customer by its purchases. But clearly the loss of a substantial customer is more important than the loss of a customer who has made a very small purchase from Mount Vernon. This difference is more accurately captured in the Secretary’s calculation.
. 19 U.S.C. § 2322(b) (1976).
. There were two types of information that supported OTAA’s and the Secretary’s conclusions regarding float glass. First, there was the trade and industry analysis that gave an overview of the flat glass industry. OTAA found that sheet glass constituted the following percentages of domestic shipments of flat glass for the stated periods:
1972: 28.5%
1974: 24.4%
1975: 16.3%
This continual decline in sheet glass producers’ market share was credited to the increasing popularity of float glass. Appendix at 13.
OTAA found that of 22 sheet glass lines operating in 1973, less than half were operating in 1976. “Since January, 1970 one of the firms producing sheet glass now produces only float glass and three of the others have built float furnaces and shut down several sheetlines.” Worker Investigation TA-W-943, id. at 11. There were only four domestic firms producing sheet glass in 1971. PPG itself now principally relies on the float process. Id. at 161-62.
A second source of data was the survey responses of Mount Vernon customers. Of the twelve customers surveyed in the first investigation, four indicated that they had decreased purchases from Mount Vernon. None had increased purchases of imports, but three indicated that they increased purchases of float glass. In the second survey, customers who expressed an opinion on Mount Vernon’s decline attributed it to the superiority of float glass and Mount Vernon’s failure to be competitive with other domestic suppliers of sheet glass in terms of quality and price.
.For this reason we find it reasonable that the OTAA and the Secretary did not go in great detail into the price suppressing effects of imports.
Petitioners speculate that “domestic prices for sheet glass may have been suppressed by the flood of low-priced foreign glass, thus reducing profits and further impairing the financial viability of domestic manufacturing facilities such as the Mt. Vernon plant.” Brief for Petitioners at 53. in fact, OTAA cited U.S. Glass Metal and Glazing (May 1976), the official publication of the Flat Glass Manufacturing Association, for the proposition that in most instances it is uneconomical to import glass. Statistics compiled by the Bureau of Labor Statistics show that the wholesale price index for window glass in 1975-76 rose almost twice as fast as the wholesale price index for all industrial commodities — hardly indicative of price suppression. U.S. Bureau of Labor Statistics, Wholesale Prices and Price Indexes (Sept.1976).
. See Groesbeck v. Duluth, S. S. & A. Ry., 250 U.S. 607, 614-15, 40 S.Ct. 38, 41, 63 L.Ed. 1167 (1919) (Brandéis, J.) (“experience teaches that it is much easier to reject formulas presented as being misleading than to find one apparently adequate”).
. Section 222 provides that a group of workers is to be certified for adjustment assistance only if “sales or production, or both, of such firm or subdivision have decreased absolutely." 19 U.S.C. § 2272(2) (1976) (emphasis added).
. 19 U.S.C. § 2273(b) (1976).
. S.Rep.No.93-1298, 93d Cong., 2d Sess. 133 (1974), reprinted in [1974] U.S.Code Cong. & Admin.News, pp. 7186, 7275.
. Id.
. See UAW v. Marshall, 189 U.S.App.D.C. 232, 584 F.2d 390 (1978).
. In explicating the “substantial” cause requirement of § 201(b), see note 12 supra, the *250House Committee on Ways and Means made the following observation:
[I]f imports were just one of many factors of equal weight, imports would meet the test of being “not less than any other cause [sic] but it would be unlikely that any of the causes would be deemed an “important” cause.
H.R.Rep.No.93-571, 93d Cong., 1st Sess. 46 (1973). As noted above, petitioners contend, and this court agrees, that there is a pertinent similarity between the “substantial cause” requirement and the “contributed importantly” provision.
. Brief for Petitioners at 23.