Alabama Ass'n of Insurance Agents v. Board of Governors of Federal Reserve System

533 F.2d 224

• Court: Court of Appeals for the Fifth Circuit
• Decided: June 10, 1976
• Docket: Nos. 74-2981, 74-3544 and 74-3843
• Status: Published

Opinion

THORNBERRY, Circuit Judge:

These consolidated petitions for review of orders by the Board of Governors of the Federal Reserve System represent the culmination of a four year campaign by Southern Bancorporation (Southern) of Birmingham, Alabama, and First National Holding Corporation (First National) of Atlanta, Georgia, to gain entry into certain aspects of the insurance agency business. From a broader perspective, they represent only one of many episodes in a huge commercial tug-of-war between the bank holding company industry on one hand and independent insurance agents and other insurance industry groups on the other. This struggle has taken place not only in the courts, but in the Congress and before the Federal Reserve System Board of Governors as well. After carefully considering the efforts and accomplishments of the contending parties before those bodies and the arguments advanced by the able counsel in this case, we find that substantial ground has been gained by the bank holding companies. We uphold for the most part the Board of Governors’ action in granting Southern and First National permission to act as agent with respect to specified types of insurance and leave further struggle between insurance agents and these holding companies in the specified areas to the marketplace.¹

I. Legislative and Procedural Background

While recognizing that the protection of savings and the extension of credit are useful and even essential to our economic wellbeing, we as a nation have often been distrustful of the concentrated economic power that may accrue to the organizations that provide those services. This feeling has not been left at large in the streets, the factories, and the small businesses of the land, but has been written into the law by Congress; the Bank Holding Company Act, 12 U.S.C. § 1841 et seq. is an expression of that sentiment.² In passing the Act in 1956, Congress sought both to limit the concentration of banking resources and to implement a policy that “bank holding companies ought not to manage or control non-banking assets having no close relationship to banking.” Sen.Rep. No. 84-1095, 84th Cong., 1st Sess. (1956), 1956 U.S.Code & Adm.News at p. 2482. In furtherance of the latter objective, acquisition by holding companies of nonbanking enterprises was prospectively forbidden. 12 U.S.C. § 1843(a) (1970). But a very important exception to the general prohibition was included: it remained acceptable to acquire companies “of a financial, fiduciary, or insurance nature . . . which the [Federal Reserve] Board . . . has determined to be so closely related to the business of banking or of managing or controlling banks as to be a proper incident thereto .” § 4(c)(8) of the Bank Holding Company Act of 1956,12 U.S.C. § 1843(c)(8) (1970).³

Pursuant to this provision, the Federal Reserve System Board of Governors (hereinafter the Board) in the late 1950’s and 1960’s granted many individual applications by holding companies to engage in non-banking activities, including the sale of insurance. And when amendments to the Act were discussed in the late 1960’s, there was significant sentiment for broadening the Board’s discretion in the area. Whether the proponents of liberalization succeeded will be dealt with hereinafter; but it is clear that the Holding Company Act Amendments passed in 1970 contained several changes of significance in this case. First, Congress permitted the Board to make the determination whether particular lines of business were closely related to banking and thus within the § 4(c)(8) exception by general regulation, as well as by order in individual cases. Second, the limitation of such activities to those “of a financial, fiduciary, or insurance nature” was deleted. Third, Congress manifested unhappiness with the language requiring § 4(c)(8) activities to be “closely related to the business of banking”; it required in 1970 only that the business be “closely related to banking”. Fourth, the Board was permitted to “differentiate between activities commenced de novo and activities commenced by the acquisition, in whole or in part, of a going concern.” Finally, and most significantly, a “public benefits test” was added to the criteria under § 4(c)(8); it provides that:

[I]n determining whether a particular activity is a proper incident to banking or managing or controlling banks the Board shall consider whether its performance by an affiliate of a holding company can reasonably be expected to produce benefits to the public, such as greater convenience, increased competition, or gains in efficiency, that outweigh possible adverse effects, such as undue concentration of resources, decreased or unfair competition, conflicts of interest, or unsound banking practices.

The Federal Reserve Board was not slow to use its new power to issue § 4(e)(8) regulations. After rulemaking proceedings in 1971, in which petitioner National Association of Insurance Agents (NAIA) participated, the Board promulgated § 225.4(a) of Regulation Y, which lists several nonbanking activities deemed sufficiently related to banking to satisfy the requirements of § 4(c)(8). Those relating to the insurance industry and to this appeal are set out in § 225.4(a)(9):

(a) . . . The following activities have been determined by the Board to be so closely related to banking or managing or controlling banks as to be a proper incident thereto: .

(9) acting as insurance agent or broker in offices at which the holding company or its subsidiaries are otherwise engaged in business (or in an office adjacent thereto) with respect to the following types of insurance:

(i) Any insurance for the holding company and its subsidiaries;

(ii) Any insurance that (a) is directly related to an extension of credit by a bank or a bank-related firm of the kind described in this regulation, or (b) is directly related to the provision of other financial services by a bank or such a bank-related firm, or (c) is otherwise sold as a matter of convenience to the purchaser, so long as the premium income from sales within this subdivision

(ii) (e) does not constitute a significant portion of the aggregate insurance premium income of the holding company from insurance sold pursuant to this subdivision (ii);

(iii) Any insurance sold in a community that (a) has a population not exceeding 5,000, or (b) the holding company demonstrates has inadequate insurance agency facilities.

12 C.F.R. § 225.4(a)(9) (1976).

No judicial review of this regulation was immediately sought. On September 6, 1972, the Board issued a regulation indicating its views as to the operative meaning of § 225.4(a)(9) in several particulars. § 225.-228 of Regulation Y, 12 C.F.R. § 225.228 (1976). NAIA sought review of that regulation in the Court of Appeals for the District of Columbia Circuit, but that court dismissed the petition, holding that the regulation was merely interpretive and did not have the force of law. Because the issues thereby raised could be litigated in connection with specific applications under the Act, the court found the petition premature. National Ass’n of Ins. Agents, Inc. v. Board of Governors of Federal Reserve System, 160 U.S.App.D.C. 144, 489 F.2d 1268 (1974).

Among the flurry of applications for permission to engage in nonbanking activities which followed the adoption of § 225.4(a) of Regulation Y were those of Southern Ban-corporation and First National Holding Corporation. They were submitted separately in February of 1972, but sought authority to sell, through subsidiaries, many of the same types of insurance. Some of the lines, such as credit life, credit health and accident insurance, and mortgage redemption insurance are not in issue here. The applications to sell other types of insurance were and are hotly contested. In this category are property damage insurance on assets being financed by banks owned by the holding company, liability insurance for borrowers from such banks, and “convenience” insurance as defined in Regulation Y.

The insurance industry parties in this case immediately expressed their opposition to the Southern and First National applications. After more than a year of preparation and preliminary skirmishing, the applications were considered consecutively at hearings before Administrative Law Judge Paul N. Pfeiffer in June, 1973. In early 1974, Judge Pfeiffer announced his findings and recommended decision: although the insurance activities proposed by Southern and First National were within the “closely related” standard of § 4(c)(8), as reasonably interpreted by the Board in its regulation, the holding companies had failed to meet the “public benefits” test added to the Act in 1970 (as to all but proprietary insurance for the holding company and its subsidiaries, and credit life, credit health and accident, and mortgage redemption insurance). In particular, the Administrative Law Judge found that there did not appear to be a reasonable expectation of substantial increased convenience to the public; that there was no basis for an expectation of operating economies which might be passed on to the consumer; that there were possibilities of unfair competition in the sale of insurance' through “voluntary tying” of insurance purchases to bank loans; that the market power of each holding company was such that it could divert certain captive clientele of the banks from existing independent insurance agencies; and that sales of surety bonds could involve the holding companies in conflicts of interest.

The holding companies were down, but by no means out. They pressed their claims vigorously before the Federal Reserve Board and, in July and September of 1974, the Board responded favorably. In its orders, the Board noted its previous determination in § 225.4(a)(9) that certain insurance activities were permissible within the meaning of § 4(c)(8) of the Bank Holding Company Act. It adhered to that determination and found that, with minor exceptions,⁴ the applications by Southern and First National were within the regulation. The Board generally accepted the basic factual findings of the Administrative Law Judge, but drew almost directly opposite conclusions from them. It concluded that it was reasonable to expect that the proposed activities would likely produce benefits to the public in convenience, efficiency, and increased competition, and that there was no danger of undue concentration of resources, no likelihood of conflicts of interest, and no evidence that there would be unfair competition in the form of coercive or voluntary tying of loans and insurance purchases. Accordingly, both Southern and First National were given permission to act as broker with respect to all of the lines of insurance presently in dispute.

II. Parties

The petition for review in No. 74-2981 was brought by the National Association of Insurance Agents, the Alabama Association of Insurance Agents, the Birmingham Association of Insurance Agents, and several individual insurance agencies. NAIA also is a petitioner in No. 74-3544, along with the Georgia Association of Independent Insurance Agents and several individual agencies. Together the 74-2981 and 74-3544 petitioners shall be referred to hereinafter as “the NAIA parties”. All petitioners are or represent individual insurance agencies which are primarily small businesses selling insurance policies issued by more than one underwriter.

Southern Bancorporation (formerly known as The Alabama Financial Group, Inc.) is a bank holding company headquartered in Birmingham, Alabama. Its primary business is banking and it has subsidiaries in that business operating in twelve Alabama cities. Several of these subsidiaries have more than one banking office; the Birmingham Trust National Bank has over twenty. Southern also has three non-banking subsidiaries: AFG Data Services, Inc., Alabama Financial Leasing, Inc., and Southern State Life Insurance Co.

Southern is the third largest bank holding company in Alabama, controlling approximately ten per cent of all commercial bank deposits in the state. It controls about twenty-two per cent of bank deposits in the Birmingham metropolitan area and nearly half of deposits in the Dothan metropolitan area and in Marion County. With the other four largest bank holding companies in Alabama, Southern controls approximately seventy-five per cent of commercial bank deposits in the state.

First National Holding Corporation, unlike Southern, is a one-bank holding company.⁵ Its banking subsidiary, First National Bank of Atlanta, has over forty offices in two counties of the Atlanta metropolitan area. It is the second largest bank in the area, with about twenty-five per cent of all bank deposits. The holding company also has 162 subsidiaries engaged in various business and financial endeavors in Georgia and neighboring states. The most important of these for our purposes is Tharpe & Brooks, Inc., a wholly owned subsidiary which is in the mortgage banking business. Unlike Southern, which proposes to establish a new insurance agency to carry on the activities, First National plans to conduct its insurance activities through First South Insurance Agency, a Tharpe & Brooks subsidiary.

A. Jurisdiction to Review

The jurisdiction of this court derives from 12 U.S.C. § 1848, which authorizes review of a Federal Reserve Board “order” at the instance of a person thereby aggrieved. On the basis of the wording of this provision, the Board asserts that we lack power to review § 225.4(a)(9) of Regulation Y, which was promulgated in 1971 and not challenged at that time. The Board correctly notes that there are significant legal differences between orders and regulations, and argues that the finality of a great number of Board orders under the regulation would be cast in doubt if review of the regulation were permitted at this time.

We disagree with the Board and find that it is our duty to review not only these orders, but the regulations upon which they are partially based. For one thing, the courts have held that a statutory provision for review of orders permits review of regulations, even standing alone, where the issues are primarily legal rather than factual, and there is an adequate record for review. See Deutsche Lufthansa Aktiengesellschaft v. CAB, 156 U.S.App.D.C. 191, 479 F.2d 912 (1973); City of Chicago v. FPC, 147 U.S.App.D.C. 312, 458 F.2d 731 (1971), cert. denied, 405 U.S. 1074, 92 S.Ct. 1495, 31 L.Ed.2d 808 (1972). See also Pacific Gas & Elec. Co. v. FPC, 164 U.S. App.D.C. 371, 506 F.2d 33, 47-48 (1974). This principle was illustrated in the recent decision in National Courier Ass’n v. Board of Governors of Federal Reserve System, 170 U.S.App.D.C. 301, 516 F.2d 1229 (1975) (Couriers), in which the court reviewed amendments to § 225.4(a) of Regulation Y under the jurisdictional provision for review of “orders” invoked here. The case before us arguably fulfills the applicable criteria, since the issues are primarily legal — whether the activities listed in the regulation meet the statutory “closely related” test— and there is factual material concerning the relationship of these activities to banking in the voluminous record compiled in these proceedings.

More importantly, we believe that even belated review of these regulations should not be foreclosed in the circumstances of this case. We deal here with orders that involve two basic issues concerning the holding company applications: whether they meet the “closely related” test and whether they meet the “public benefits” test. The latter, by its nature, may be conclusively resolved only in the context of specific applications,⁶ Independent Bankers Ass’n of Georgia v. Board of Governors of Federal Reserve System, 170 U.S.App.D.C. 278, 516 F.2d 1206, 1215-16 (1975); American Bancorp., Inc. v. Board of Governors of Federal Reserve System, 509 F.2d 29, 38-39 (8 Cir. 1974), while the former can be resolved either on an individual basis or by regulation. See 12 U.S.C. § 1843(c)(8). Although the Board’s decision to proceed by regulation was clearly proper, we note that its effect is not merely to narrow the area of contention between the parties on the “closely related” question, but in some respects to conclude that question in favor of the applicants.⁷

In light of the determinative effect of the regulation on some issues involved in the promulgation of these orders, it would be an “empty and useless thing” to review them without scrutiny of the regulation as well. Doe v. CAB, 356 F.2d 699, 701 (10 Cir. 1966), quoted in International Harvester Co. v. Ruckelshaus, 155 U.S.App.D.C. 411, 478 F.2d 615 (1973). We would partially abdicate our judicial duty if we were to allow the time limit for direct review to insulate forever this regulation, which has never been judicially tested, from scrutiny as to its consistency with Congressional mandate, all the while “reviewing” orders based, in significant part, wholly and directly upon it. Perhaps more importantly, that course would be unjust to many of the petitioning parties. It may be fair to say that NAIA, which participated in the rule-making proceedings in 1971, had a chance to challenge this regulation within thirty days of its promulgation.⁸ But is it fair to say the same thing of petitioner R. A. Brown Insurance Agency, Inc., in Alabama, or Morris-Fallaize Insurance, Inc., in Georgia, which may bear the economic detriments potentially arising from these applications? We believe not and find that, considerations of finality notwithstanding, our duty to see that the Congressional mandate in § 4(c)(8) is carried out in the individual cases which come before us requires review of § 225.4(a)(9) of Regulation Y. See Oljato Chapter of Navajo Tribe v. Train, 169 U.S.App.D.C. 195, 515 F.2d 654, 659 & n. 6 (1975); Functional Music Inc. v. FCC, 107 U.S.App.D.C. 34, 274 F.2d 543 (1958), cert. denied, 361 U.S. 813, 80 S.Ct. 50, 4 L.Ed.2d 81 (1959).

B. Procedural Validity

We pass then to the question of whether the regulation complied with the requirement of the Administrative Procedure Act that

the agency shall incorporate in the rules adopted a concise general statement of their basis and purpose.

5 U.S.C. § 553(c) (1970). In this case, the regulation contained nothing that might be thought to respond to this statutory command other than the statement that

[t]he following activities have been determined by the Board to be so closely related to banking or managing or controlling banks as to be a proper incident thereto.

This statement does no more than conclusorily announce the import of the regulation. It does nothing to illuminate the process by which the Board arrived at the regulation and accordingly is of little use in fulfilling the purposes of the statute, the facilitation of judicial review. See Amoco Oil Co. v. EPA, 163 U.S.App.D.C. 162, 501 F.2d 722 (1974); Automotive Parts § Accessories Ass’n v. Boyd, 132 U.S.App.D.C. 200, 407 F.2d 330 (1968); National Resources Defense Council, Inc. v. SEC, 389 F.Supp. 689 (D.D.C.1974). The APA provision “does not require the agency to supply specific and detailed findings and conclusions of the kind customarily associated with formal proceedings”, but this limitation “does not lessen or excuse the agency’s obligation to publish a statement of reasons that will be sufficiently detailed to permit judicial review.” National Nutritional Foods Ass’n v. Weinberger, 512 F.2d 688, 701 (2 Cir. 1975).

This statutory requirement, however, has not been enforced with the vigor that would, at first blush, seem indicated. Statements with no more informative content than the one before us have been held to pass muster. Automotive Parts & Accessories Ass’n v. Boyd, supra; New York Freight Forw’rs & Brokers Ass’n v. Federal Maritime Comm’n, 337 F.2d 289 (2 Cir. 1964), cert. denied, 380 U.S. 910, 85 S.Ct. 893,13 L.Ed.2d 797 (1965). And regulations with no statement of basis and purpose at all have been upheld where the court deemed the basis and purpose obvious. Hoving Corp. v. FTC, 290 F.2d 803 (2 Cir. 1961).

The rationale of these decisions is not difficult to ascertain. Courts must have an adequate basis to engage in judicial review, but, as partners with the agencies in the effectuation of Congressional will through the administrative process, they do not function to strike down agency action because of merely formal or technical flaws. Greater Boston Tel. Corp. v. FCC, 143 U.S. App.D.C. 383, 444 F.2d 841, 851 (1970), cert. denied, 403 U.S. 923, 91 S.Ct. 2229, 29 L.Ed.2d 701 (1971); Braniff Airways v. CAB, 126 U.S.App.D.C. 399, 379 F.2d 453, 465-68 (1967). For this reason, a court must not only examine whether an agency’s promulgation of a challenged regulation complies with the procedural requirement; it must also determine whether, in light of the nature and content of the regulation and of the underlying legislation, the extraneous material which may be available to explain the basis and purpose of the agency action, and the quantum of action taken in reliance on the regulation, any procedural flaw so subverts the process of judicial review that invalidation of the regulation is warranted.

We find that this case is one in which we can “discern the path” of the Board with sufficient confidence. See Environmental Defense Fund, Inc. v. EPA, 167 U.S.App.D.C. 71, 510 F.2d 1292, 1304 (1975), quoting Greater Boston Television Corp. v. FGC, 143 U.S.App.D.C. 383, 444 F.2d 841, 851 (1970), cert. denied, 403 U.S. 923, 91 S.Ct. 2229, 29 L.Ed.2d 701 (1971). The regulation at issue is not one which purports to carry out a vague statutory command through prescription of detailed standards of conduct, or sets out a complex method of proceeding in order to receive some type of agency approval. Cf. National Ornament & Elec. Light Christmas Ass’n, Inc. v. Consumer Product Safety Comm., 526 F.2d 1368, 1371-72 (2 Cir. 1975). To the contrary, the regulation makes a single and limited factual and legal determination: that certain insurance activities are so “closely related” to banking as to be proper incidents thereto. For this reason, it is not necessary for the agency to explain what the purpose of the regulation is; it obviously is intended to facilitate the process of applying for holding company authority to engage in non-banking activities while authorizing only such activities as are permissible under the “closely related” standard.

The basis for the agency’s designation of these particular activities as permissible is less clear. But to aid us in divining it we have not only the detailed briefs of the several parties and amici, the pre-1970 decisions of the Board on applications to engage in insurance activities,⁹ and the legislative history of the 1956 Act and the 1970 Amendments, but also the extensive record in the hearings before the Administrative Law Judge in these cases. Cf. Automobile Parts & Accessories Ass’n v. Boyd, supra at 338 (court may consider statement of agency in denying rehearing); Environmental Defense Fund v. EPA, 167 U.S.App.D.C. 71, 510 F.2d 1292, 1304 & n. 40 (1975) (resort to detailed factual record is permissible in lieu of complete findings in hearing contest). Admittedly, much of the record is not relevant to the Board’s reasoning on the “closely related” issue and that portion which is relevant is “after the fact” explication of which we have every right to be suspicious. See Rodway v. Department of Agriculture, 168 U.S.App.D.C. 387, 514 F.2d 809, 816-17 (1975). But we find that there is enough here, in light of the limited nature of the regulation and the impact of its wholesale invalidation for non-substantive flaws, that we can “see whether the agency, given an essentially legislative task to perform, has carried it out in a manner calculated to negate the dangers of arbitrariness and irrationality.” Automotive Parts & Accessories Ass’n v. Boyd, supra at 338.

C. Substantive Challenges

The NAIA parties generally argue that the activities listed in § 225.4(a)(9) of Regulation Y violate § 4(c)(8) of the Bank Holding Company Act in that they are not sufficiently “closely related to banking.” Although their efforts are directed most vigorously toward the portion of the regulation which the Board has relied upon in authorizing the sale of property and liability insurance in these orders, we find that the very general language of the regulation on that subject is permissible. Other portions, however, authorize the sale of insurance that is so clearly unrelated to banking that they must fall.

As the Board recognizes, the “closely related” standard imposes definite limits upon its discretion to permit nonbanking activities by bank holding companies. Accordingly, the Board rejected applications to engage in some insurance activities, even before the 1970 Amendments.¹⁰ However, the great majority of pre-1970 insurance applications were approved by the Board, and these constituted most of the nonbanking activities authorized under § 4(c)(8) pri- or to 1970. Among the specific types of activities authorized were insurance for the holding company and its employees, comprehensive property insurance, and renewal of insurance covering collateral on a loan after it had been repaid.

It was against this backdrop that Congress acted in 1970. Both the Senate and House of Representatives passed legislation which was intended in part to give the Board greater discretion in administering the Act.¹¹ In particular, both houses approved a change of the “closely related” language of the 1956 Act to a requirement that nonbanking activities need be only “functionally related” to banking.¹² The two bills did not completely agree, however; the House of Representatives included a provision specifically prohibiting holding companies from engaging in six nonbanking activities, including insurance agency activities other than those relating to credit life, and credit health and accident insurance.

The conflict between the “functionally related” language, which clearly was intended to broaden to some degree the Board’s discretion in approving nonbanking activities, and the specific prohibitions in the House bill made for a difficult Conference Committee session. Ultimately, the list of prohibited activities was deleted from the Act, apparently in return for a retention of the pre-1970 “closely related” language. Four of the seven House conferees ¹³ emerged to argue that the resulting legislation should “be construed to mean that the Congress was not convinced that such expansion and liberalization was justified.” Conference Report, 1970 U.S.Code Cong. & Adm.News at p. 5572. See also 116 Cong.Rec. 41952, 41961 (1970) (remarks of Rep. Patman). The Senate conferees and the minority House conferees, however, argued that even the compromise bill which survived the Conference represented a liberalization of the law.¹⁴

Certainly some other changes in the Act represent a modest broadening of the Board’s discretion. The deletion of the requirement that the nonbanking activity be of a “financial, fiduciary, or insurance nature” and that it relate to the “business of banking” are examples.¹⁵ And it can be argued that the belief of the majority of the entire Conference Committee should be given precedence over that of the slim majority of the House conferees. Nevertheless, on balance we cannot concur in the District of Columbia Court of Appeal’s conclusion that “it is plain that Congress intended to expand [the Board’s] latitude in substantial degree”. 516 F.2d at 1236. Any liberalization of the standards effected by these minor wording changes was largely if not wholly offset by the Conference Committee’s rejection of the “functionally related standard” and the addition of the public benefits test. The purpose and effect of the 1972 Amendments in this particular was quite ambiguous, and we find that no significant change in the general range of the Board’s discretion was effected by them.

Despite petitioners’ earnest efforts to persuade us to this conclusion, however, we find that it avails them little. One reason for this is that the pre-1970 law simply was not very restrictive in the insurance area. Petitioners correctly note that approval of many pre-1970 insurance applications was grounded in part on long established bank practices in the local area involved and the existence of bank competitors already carrying on insurance activities.¹⁶ But, even so, the fact remains that the Board between 1956 and 1970 rejected the first two applications to engage in insurance activities and thereafter granted all twenty-five applications which came before it, many without reference to local practices; this hardly bespeaks a highly restrictive policy.

Petitioners have been able to cite nothing in the legislative history to indicate disapproval of the Board’s pre-1970 decisions. To the contrary, the Senate Report stated that

the deletion of [the restriction to “financial, fiduciary, or insurance” activities] shall not be construed as indicating the committee’s intent that these activities be no longer permitted. To the contrary, these activities are “functionally related to banking” and are included under the new provision adopted by the committee to the full extent they were permitted under the existing act, and to such additional extent as the Board may find it desirable under the new criteria established.

Senate Report, 1970 U.S.Code Cong. & Adm.News at p. 5533 (emphasis supplied). Of course, the “functionally related” language disappeared in conference, but even the House members who held the more restrictive view of the resulting legislation quoted the Senate Report’s language, emphasizing that “these . . . are included . . .to the full extent they were permitted under the existing act.” Conference Report, 1970 U.S.Code Cong. & Adm. News at p. 5565. See also 116 Cong.Rec. 41963 (remarks of Rep. Reuss).

We have, then, a series of interpretations of the “closely related” standard with respect to insurance activities, over a period of fourteen years, by the agency charged with setting the machinery of the Bank Holding Company Act in motion. See Chemehuevi Tribe of Indians v. FPC, 420 U.S. 395, 409-10, 95 S.Ct. 1066, 1075, 43 L.Ed.2d 279, 290 (1975); Trafficante v. Metropolitan Life Ins. Co., 409 U.S. 205, 210, 93 S.Ct. 364, 367, 34 L.Ed.2d 415, 420 (1972); FDIC v. Sumner Fin. Corp., 451 F.2d 898 (5 Cir. 1971). The “great respect” that is due such interpretations “is enhanced by the fact that Congress gave no indication of its dissatisfaction of the agency’s interpretation of the scope of its . jurisdiction when it amended the Act.” Chemehuevi Tribe of Indians, supra. See Saxbe v. Bustos, 419 U.S. 65, 95 S.Ct. 272, 42 L.Ed.2d 231 (1974); Reeb v. Economic Opp’y Atlanta, Inc., 516 F.2d 924 (5 Cir. 1975). We must be most reluctant then, to reach results which would repudiate pre-1970 Board decisions, particularly in view of the fact that the Board’s expertise in this area must be presumed to exceed ours. See Brennan v. General Tel. Co., 488 F.2d 157, 160 (5 Cir. 1973).

A second reason why petitioners have a difficult burden derives from our limited role in reviewing these regulations. Because this regulation does not involve findings “required by statute to be made on the record after opportunity for an agency hearing,” see 5 U.S.C. §§ 553, 556-57 (1970); Couriers, 516 F.2d at 1235 n. 6, the “substantial evidence” standard of review is not the proper one. See United States v. Allegheny-Ludlum Steel Corp., 406 U.S. 742, 749, 92 S.Ct. 1941, 1946-47, 32 L.Ed.2d 453, 460 (1972); Couriers, 516 F.2d at 1235 n. 7. Instead, we need determine only whether the Board’s findings of fact are “arbitrary, capricious, an abuse of discretion, or otherwise -not in accordance of law”, 5 U.S.C. § 706 (1970), and determine questions of law for ourselves. Couriers, 516 F.2d at 1235-36.

With these observations we turn to the portion of § 225.4(a)(9) under primary attack. It provides that a holding company subsidiary may act as an insurance agent or broker with respect to:

(ii) Any insurance that (a) is directly related to an extension of credit by a bank or a bank-related firm of the kind described in this regulation, or (b) is directly related to the provision of other financial services by a bank or such a bank-related firm. .

The immediately apparent fact is that these sections represent only slightly more narrow definitions than the “closely related” language itself. They state in essence that one type of “close” relation to banking exists when there is a “direct” relationship of the activity to an extension of credit or other financial service. Since the essential function of banking is the provision of credit or other financial services, the primary result of the regulation itself is merely to shift the inquiry from the word “close” to “direct”. Such a minute change cannot be deemed outside the Board’s discretion; if there is invalidity in the Board’s action pursuant to this regulation, it must be in the application of the regulation to the particular activities authorized under it.

Other parts of § 225.4(a)(9) represent more substantial departures from the wording of the Act. The first example is subsection (i), which permits the sale of “any insurance for the holding company or its subsidiaries”.¹⁷ Acting pursuant to this authorization, the Board approved applications by both Southern and First National and authorized the sale of any insurance within the terms of subsection (i).

It is apparent that at least some of the insurance sales thereby permitted will relate only tangentially to the basic functions of banks: the protection of savings and the provision of financial services. For example, the regulation would authorize the sale of group health insurance for janitorial and maintenance employees, and property insurance on the physical assets of banking subsidiaries, such as bank fixtures or real property owned by the bank. Moreover, sales of insurance to the holding company or to nonbanking subsidiaries would be permitted. In the case of First National, for instance, the insurance subsidiary could sell insurance to Gulf Finance Corporation of Mississippi or Woods-Tucker Leasing Company, which has offices in Mississippi and Louisiana. However, convenient to the holding company it is to eliminate the need to pay the commissions of an unrelated insurance broker for such insurance, the relationship of the brokering of such insurance to banking is difficult to discern.

We do not seek to be restrictive in determining what types of relationships to banking are cognizable under the Act. As the Couriers court stated, activities meeting the following criteria seem permissible:

1. Banks generally have in fact provided the proposed services.

2. Banks generally provide services that are operationally or functionally so similar to the proposed services as to equip them particularly well to provide the proposed service.

3. Banks generally provide services that are so integrally related to the proposed services as to require their provision in a specialized form.

516 F.2d at 1237. Clearly, however, brokering insurance for even those subsidiaries of a holding company which are banks does not come within these strictures. We have been apprised of no general practice on the part of banks (and certainly the Board has not articulated one) to broker the types of insurance that they need for their own account and we know of no evidence that they have provided “functionally” or “integrally” related insurance services. Certainly any past practice in which banks have engaged in this area does not cover “any” type of insurance which the banks or their subsidiaries might require.

We are willing, however, to consider broader types of relationships. For example, it cannot be doubted that many types of insurance for banking subsidiaries are helpful, and perhaps essential, to the success of the enterprise, and thus to the provision of financial services. Such insurance may protect the physical facilities which are necessary to banking, insure the fidelity of the employees who are intimately involved in banking operations, or insure the availability of necessary health care for the employees who directly or indirectly provide financial services. The relationship of such insurance to “banking” is in some cases quite indirect; in many cases the bank’s need for it is no different than the need of any business enterprise. But in view of the limited role of this court in reviewing these regulations, we are unwilling to say that the Board was wrong in finding it sufficient.

As noted above, however, even this rationale cannot justify the Board’s authorization of the holding companies to broker insurance for themselves or their nonbanking subsidiaries. Such insurance simply does not contribute to the operations of those subsidiaries actually engaged in the banking business. Perhaps it could be argued that uninsured losses of those businesses or of the holding company may affect the availability of resources to the banking subsidiary or other policies of the directors or shareholders of the holding company, who directly or indirectly control the banking subsidiaries. But the Board has not, to our knowledge, articulated this rationale and, in any event, it seems to us so attenuated as to stray even from the broad area of discretion to which the Board is entitled. We therefore find that the portion of § 225.4(a)(9)(i) which authorizes the sale of insurance for bank holding companies and nonbank subsidiaries of holding companies must be invalidated, and the portions of the orders under review which authorize the sale of such insurance must also fall.

A second specific provision in the regulation is (ii)(e), which authorizes holding company subsidiaries to act as agents with respect to insurance that

is otherwise sold as a matter of convenience to the purchaser, so long as the premium income from sales within this subdivision (ii)(c) does not constitute a significant portion of the aggregate insurance premium income of the holding company from insurance sold pursuant to this subdivision (ii). .

The Board argues that this provision “permits the holding company agency to round-out its insurance offerings” by, for example, permitting renewal of policies covering an automobile after a loan from a holding company bank on the automobile has been repaid. The Board stresses also that, in its interpretive regulation, 12 C.F.R. § 225.-128(e), it limited premiums from sales of “convenience insurance” to less than five per cent of the aggregate premium income of the holding company agency, and stated specifically that this provision was not designed to permit entry into a general insurance agency business. It further argues that, as the business of a holding company agency increases over time, a larger portion of its convenience insurance sales will be renewals of insurance on property on which indebtedness has been fully repaid, until that is the only type of convenience insurance sold.

The Board may well be correct in its forecasts. And it may be correct in assuming that such renewal insurance would be “closely related” to banking within the meaning of the Act. But the fact is that the regulation is not so restricted. It permits holding company affiliates to sell any type of insurance under any conditions as a “matter of convenience”, imposing no condition that the insurance relate in any way to banking.

The court in the Couriers case was faced with a similar provision, which authorized holding company affiliates to furnish “courier services for non-financially-related material upon the specific, unsolicited request of a third party when courier services are not otherwise reasonably available.” 12 C.F.R. § 225.129 (1974). Although the court was willing to agree that an affiliate could engage in such incidental activities as were reasonably necessary to carrying out those that were closely related to banking, it stated that:

The justification for the “incidental” courier services at issue here is, however, quite different. It is not, as far as we can tell, that the carriage of non-financially related material is in any way necessary to the successful operation of the courier service affiliates. Rather, it is that the provision of such service would “serve the convenience of the public”.

516 F.2d at 1240.

The court then noted that the provision of such services might indeed be beneficial to the public, in view of the fact that bank affiliates would not likely become deeply involved in providing such unsolicited services, and that no bank competitor would be injured since the services were required to be “otherwise unavailable”. The court then made a statement with which we fully agree:

But Congress did not instruct the Board to allow or disallow bank involvement in non-financial activities as may be required by the policies which counsel a separation of commerce and banking. Having heeded that counsel itself, it decreed the separation, and instructed the Board to enforce it.

516 F.2d at 1241.

Here, too, we believe the Board has misconstrued its role. The Board’s function is not to promote generally the public interest consistent with the policies of the Bank Holding Company Act, but to determine what activities are “closely related” to banking. While we might agree that the sale of certain types of convenience insurance would help a holding company agency “round out” its offerings to the benefit of its clients, the agency, and the general public, the fact remains that the handling of convenience insurance simply is not, as broadly defined in the regulation, closely related or necessarily related at all to banking.

The Board, however, argues that the brokering of convenience insurance is “a necessary activity to the operation of an affiliated insurance agency”. To the extent that this is an assertion that it is an incidental activity reasonably necessary to the carrying on of clearly permissible ones, we find that it is one that, if supported, would be legally sufficient. See 12 C.F.R. § 225.-4(a) (1976); cf. 12 U.S.C. § 24 (1970) (authorizing national banks “to exercise . all such incidental powers as shall be necessary to carry on the business of banking”). But we see no possible argument that the regulation, as written, meets this standard in view of its failure to restrict the types of insurance authorized. Moreover, as the Board itself argues, the provision would enable an agency to meet “peculiar additional insurance needs of a few of its customers”; it is difficult to see how such a power could be deemed “necessary” to the agency’s operation. Although the Board may redraft § 225.4(a)(9)(ii)(e) in a way that meets the “closely related” standard, we must hold that provision as it stands invalid and reverse those portions of the orders in question which authorize the sale of convenience insurance.

A final provision which must be examined is § 225.4(a)(9)(iii) of Regulation Y, which permits a holding company agency to sell:

Any insurance sold in a community that (a) has a population not exceeding 5,000, or (b) the holding company demonstrates has inadequate insurance agency facilities.

This provision, too, seems to authorize the handling of insurance for the benefit of the public, despite the absence of any discernible connection to “banking”. It is interesting to note, however, that Congress has since 1916 permitted national banks to engage in just such activity. 12 U.S.C. § 92 (1970). The question then arises whether the Congressional determination that such activity is permissible for national banks alone is sufficient to render it “closely related to banking” for purposes of the Bank Holding Company Act.

The answer to this question can be found in the legislative history of 12 U.S.C. § 92, which this court examined in Saxon v. Georgia Ass’n of Ind. Ins. Agents, Inc., 399 F.2d 1010 (5 Cir. 1968). In that case, we placed emphasis on a letter from the Comptroller of the Currency to the Senate Banking and Currency Committee in 1916 which revealed that his primary purpose in proposing the legislation was to give small town banks, which had difficulty in deriving a sufficient profit from banking, an additional source of revenue. The Comptroller stated that:

This additional income will strengthen them and increase their ability to make a fair return to their shareholders, while the new business is not likely to assume such proportions as to distract the officers of the bank from the principal business of banking.

399 F.2d at 1015. It clearly appears, then,' that 12 U.S.C. § 92 was enacted not out of a belief that there was any significant connection between banking and the sale of insurance in small communities, but merely because it was believed that banks needed an additional source of income to improve their stability and profitability. Congress thus did not indicate that the sale of insurance in small towns was part of “banking”, but rather that it was an unrelated business which would provide a source of income that would then be available for use in the separate banking activities of small town national banks.

No one doubts that subsidiaries of holding companies which are national banks located in small towns have the authority to broker insurance. But to hold that other holding company subsidiaries which are not banks have such authority as a result of 12 U.S.C. § 92 would not only violate the Bank Holding Company Act, but would warp the Congressional intention underlying the 1916 enactment as well. Whatever benefits in terms of convenience to the public or profit to holding companies may accrue as the result of granting affiliates power to sell insurance in small towns, it can hardly be argued that such authority is necessary to protect the stability of the banking (or other) subsidiaries of holding companies or to permit a fair rate of return to their shareholders. The purpose of Congress in enacting 12 U.S.C. § 92 simply does not apply to the bank holding companies, and that provision cannot serve as the sole authorization for a determination that insurance sales in small towns are “closely related to banking”.

This is made even more apparent by the fact that the Board regulation is not even restricted to holding company subsidiaries which are located in small communities. Unlike the national banks which were the object of the 1916 provision, the holding company affiliates which would sell this insurance would be located in downtown Birmingham or Atlanta and have full access to the commercial opportunities of major metropolitan areas and beyond. For all of these reasons, we must hold that § 225.-4(a)(9)(iii) of Regulation Y is invalid and that the portions of the orders before us which are based upon it must also fall.

III. Application of the Regulation

We turn now to one of the most hotly contested aspects of this case: the Board’s authorization of holding company sales of property damage and liability insurance. The “property damage” insurance referred to would protect collateral for loans by holding company banks from damage caused by fires, storms, accidents, and other hazards. The liability insurance at issue would protect borrowers from liability resulting from injury or damage to the person or property of others in connection with property financed by holding company banks. Thus, for example, a bank financing and taking a security interest in an automobile might also sell the owner-borrower collision insurance and insurance against any liability arising from an accident in the automobile. The Board found that both property damage and liability insurance were “directly related to an extension of credit” and therefore within § 225.-4(a)(9)(ii)(a) of Regulation Y.

Petitioners argue primarily that the sale of physical damage insurance is unauthorized because it is neither “functionally equivalent to an extension of credit” nor “operationally integrated into the lending transaction”. As noted earlier, we are not convinced that Congress intended to so limit the universe of relevant connections between banking and proposed nonbanking activities. To the contrary, we believe that the Board should be upheld when it articulates any type connection between banking and the nonbanking activity in question which makes it rational to consider the proposed activity an “incident” to banking.

In approving the sale of property damage insurance, the Board stressed the fact that such insurance is “essential from the lender’s standpoint to assure that the value of the collateral will not be impaired by physical damage.” II App. 187; IV App. 732. The Board also noted that such insurance performs “an integral function for the borrower as well, since the presence or lack of insurance protecting loan collateral is an essential element of the credit evaluation.” II App. 187. Its decision, then, was based neither on any historical practice by banks, nor upon any process characteristic of lending transactions which would necessitate specialized forms of the activity at issue. The Board seems to have relied simply on the fact that, in lending transactions, banks have a legitimate need for physical damage insurance to protect the value of the collateral and that borrowers have need for the insurance to obtain credit. We hold that this is enough, particularly in view of the pre-1970 authorization of similar insurance sales. See Citizens and Southern Holding Co., 1969 Fed.Res.Bull. 673; Otto Bremer Co., 1969 Fed.Res.Bull. 388; Otto Bremer Co., 1967 Fed.Res.Bull. 1555. Where a product or service is regularly desired by both parties and supports the transaction in some significant way, we cannot say that the Board acts irrationally or arbitrarily in deeming the furnishing of the product or service “closely related to banking” or “directly related to an extension of credit”.

The case for liability insurance proceeds upon the same general lines but is significantly weaker. Liability insurance protects the borrower’s ability to repay a loan, which could otherwise be destroyed by an accident or other catastrophe. It therefore fulfills a need of both the borrower and lender, particularly where the borrower is an individual or a small business with limited assets. Petitioners note, however, that liability insurance, unlike physical damage insurance, is not required of borrowers as a general banking practice. And the authorization at issue is in no way limited to eases where the borrower’s ability to repay could be significantly affected by an uninsured liability.

This is not surprising in view of the fact that the Board’s orders in these cases relied on an entirely different rationale. The Board stressed the insurance industry practice of selling liability insurance in conjunction with property damage insurance protecting an automobile or other property. It concluded that

a “packaged” insurance policy, combining liability insurance with insurance relating to physical damage on property purchased from loan proceeds, fulfills a legitimate need of the lender and borrower alike at the time a loan is made. Moreover, in the case of homeowner’s insurance, it appears that it would not be economical for a borrower to procure separately the various coverages customarily packaged in such a policy.

11 App. 188; see IV App. 733. The Board’s rationale therefore comes down to this: (1) borrowers, for convenience and, in some cases, cost reasons, want liability insurance packaged with property damage insurance, which itself is incident to banking; (2) the need for liability insurance arises out of the lending transaction; (3) liability insurance to some extent protects the borrower’s ability to repay the loan.

We find that these form a sufficient basis for the brokering of liability insurance by bank holding companies. The evidence amply supports the conclusion that, from the consumer’s point of view, packaged property damage and liability policies are more desirable than the same policies separately sold. See II App. 273, 278, 309-10, 411-12; V App. 1228. To permit banks to sell only property damage insurance, then, could have the effect of substantially limiting the gains in convenience and in increased competition which Congress sought to promote. The sale of packaged liability insurance therefore may be seen as an activity “incidental” to the sale of property damage insurance and necessary for the effectuation of the Congressional purpose with respect to the latter activity. See 12 C.F.R. § 225.4(a) (1976); Couriers, 516 F.2d at 1240.

It cannot be said, moreover, that the relationship of banking to liability insurance itself is chimerical. Although banks do not customarily require it, liability insurance does in a significant way “support the lending transaction”. Indeed, assurance of the borrower’s ability to repay the loan may be more important than protection of the collateral; repayment rather than repossession is the bank’s primary objective. And, too, the need for the insurance does arise directly from the purchase of property, which is the very purpose of the loan. In attempting to answer the question whether these relationships are sufficient to establish the connection that Congress required, we have been impressed by the pre-Amendment decisions of the Board. In two cases decided in the year prior to the passage in 1970 of the provision we are construing, the Board granted applications authorizing the sale of liability insurance by a bank holding company agency, without any requirement that it be packaged with other insurance. United Virginia Bankshares, Inc., 1970 Fed.Res.Bull. 599, 600; First Security Corp., 1969 Fed.Res. Bull. 667, 668. Yet Congress made no substantial change in the language of the relevant provision and the legislative history is void of any suggestion that the pre-Amendment decisions were disapproved. In view of the familiar presumption that the legislature is aware of existing constructions of statutes which it reenacts,¹⁸ this court has every reason to believe that the more limited authorization of liability insurance in these cases was within the intent of Congress. '

IV. The Calculus of Public Benefits

The 1970 Holding Company Act Amendments transformed the general question of whether the proposed activity would be a “proper incident” to banking, in part at least,¹⁹ into the more particularized inquiry of whether the activity “can reasonably be expected to produce benefits to the public . that outweigh possible adverse effects”. 12 U.S.C. § 1843(c)(8) (1970). Nonexhaustive examples of such benefits (greater convenience, increased competition, and gains in efficiency) and adverse effects (undue concentration of resources, decreased or unfair competition, conflicts of interest, and unsound banking practices) were listed by Congress in the statute. In this case, the Federal Reserve Board made specific findings as to each of these factors and, overruling the Administrative Law Judge, held that each of the possible benefits were likely to be realized if the applications were granted and that none of the possible adverse effects were likely to occur. The Board thus concluded that overall public benefits very clearly outweighed possible adverse effects.

Congress in the 1956 Act clearly circumscribed our role in reviewing the Board’s findings: “[t]he findings of the Board as to the facts, if supported by substantial evidence, shall be conclusive”. 12 U.S.C. § 1848. If such support exists in the record, it is our duty to affirm even though we might come to a diametrically opposite conclusion on our own. North Hills Bank v. Board of Governors of Federal Reserve System, 506 F.2d 623 (8 Cir. 1974); Comm’l Nat’l Bank of Little Rock v. Board of Governors of Federal Reserve System, 451 F.2d 86 (8 Cir. 1971); First Wisconsin Bankshares Corp. v. Board of Governors of Federal Reserve System, 325 F.2d 946 (7 Cir. 1963). This is particularly true with respect to matters relating to banking industry structure and operations, areas in which the Board may be presumed to have expertise. Commercial Nat’l Bank supra; Northwest Bancorporation v. Board of Governors of Federal Reserve System, 303 F.2d 832 (8 Cir. 1962).

The reasons for deference to the Board’s reasoned judgment in the general run of eases are magnified in the context of the “public benefits” determination of § 4(c)(8). Congress required only that public benefits “reasonably” be expected to outweigh adverse effects. In so doing, it indicated recognition of the fact that the judgments required of the Board are necessarily imprecise and to some degree speculative. Not only do they relate to complex future events — often unprecedented ones— but they concern aggregate rather than discrete effects. Thus, the Board must determine not whether there will be one or more aspects of insurance agency activity by holding company affiliates that will tend to produce efficiency gains, but whether, overall, such aspects will outweigh possible sources of efficiency loss sufficiently to produce net gains. The latter determination is much more difficult than the former, and is one which is much less susceptible to detailed evidentiary support. Unless we are to impose insurmountable barriers upon those who seek to take advantage of the § 4(c)(8) authorization, which was reaffirmed by Congress in 1970, our review must be sensitive to these circumstances. See, e. g., Bowman Transp., Inc. v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 292-94, 95 S.Ct. 438, 446, 42 L.Ed.2d 447, 460 (1974).

We note, however, that the Board’s rejection of the Administrative Law Judge’s conclusions cuts the other way. While the Board is in no way bound by those conclusions, e. g., Bowman Transp., Inc. v. Arkansas-Best Freight System, Inc., supra, 419 U.S. at 297-98, 95 S.Ct. at 447-48, 42 L.Ed.2d at 462-63, “evidence supporting a conclusion may be less substantial when an impartial, experienced examiner who has observed the witnesses and lived with the case has drawn conclusions different from the Board’s than when he has reached the same conclusion”. Universal Camera v. NLRB, 340 U.S. 474, 496, 71 S.Ct. 456, 469, 95 L.Ed. 456, 472 (1947). Although we are inclined to uphold the Board where it clearly explains its reasons for differing with the Administrative Law Judge (particularly on issues upon which it may be presumed to have particular expertise), its findings will be vulnerable where it merely announces an opposite conclusion from the Administrative Law Judge without explanation. See Greater Boston Television Corp. v. FCC, 143 U.S.App.D.C. 383, 444 F.2d 841, 853 (1970), cert. denied, 403 U.S. 923, 91 S.Ct. 2229, 29 L.Ed.2d 701 (1971).

A. Greater Convenience

The Administrative Law Judge found no likelihood of increased convenience in the Southern case and a likelihood of only “minimum” increased convenience in First National’s. The Board disagreed in both instances, relying primarily on its belief that the “one-stop shopping” that would be made possible would save consumers an extra trip to an insurance agent and would avoid the necessity of obtaining the same information from the consumer twice — once for the loan and once for insurance. It noted that “borrowers have often requested insurance from Applicant’s banks in the past” in support of its contention that there is a demand for the postulated added convenience. The Board now argues also that the consumer would thereby be assured that he has insurance coverage satisfactory to the lender and that insurance and loan payments could be combined in a single bill, resulting in further consumer convenience.

As the NAIA parties argue, the Board’s conclusions on this issue are cast in doubt by the fact that the licensed insurance agents in both cases will be located at central offices in Birmingham or Atlanta. The result is that, although loan officers around the area.or state can telephone such agents to make insurance arrangements at the time loans are made, any dealings by borrower-insureds living outside those metropolitan areas with the agents thereafter will be on an impersonal and long-distance basis. Relationships with local independent agents, on the other hand, can be conducted on a more personal basis and face-to-face discussion of insurance matters will be much easier to arrange. The Board showed no recognition whatever of the resulting decrease in convenience.

Moreover, the Board made no attempt to explain why non-affiliated insurance agents could not effect the same conveniences through a one-stop shopping system. Loan officers could (and if there is a genuine bank need to insure that acceptable coverage be secured, they will) suggest to the prospective loan customer a reputable local agent; if the customer is willing, that agent could be telephoned and could perform the same functions as the proposed holding company affiliate agent.

In its brief, First National suggests that a practice of recommending independent insurance agents to borrowers would not be “good business policy or good ethics”.²⁰ This may or may not be true, but what is clear is that the Board made no such argument and that we have been directed to no evidentiary support in the record for that conclusion. The same situation exists with respect to the claim of added billing convenience; it is one advanced post hoc by the Board with no evidentiary support which has been pointed out to us. See Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168-69, 83 , 9 L.Ed.2d 207, 216 (1962); City of Lafayette v. SEC, 147 U.S.App.D.C. 98, 454 F.2d 941 (1971), aff’d sub nom. Gulf States Utilities Go. v. FPC, 411 U.S. 747, 93 S.Ct. 1870, 36 L.Ed.2d 635 (1973). Because of these deficiencies, and particularly because of the Board’s failure to delineate clearly the sources of its disagreement with the conclusions of the Administrative Law Judge, we must find that its conclusion on this point is not supported by substantial evidence.

B. Greater Efficiency

The dispute on this point centers around the insurance industry practice of “direct billing”, by which insurance underwriters, rather than agents, issue the policy to the customer and bill and collect the premiums. There was evidence that this practice results in lower commissions to the agent and lower premiums for the consumer. Because both First National and Southern indicated that they probably would not use “direct billing”, the Administrative Law Judge concluded that any possible gains in efficiency from elimination of, advertising and soliciting expense and from combination of billing would be negated, and that therefore there was no reasonable likelihood of efficiency gains.

The Board, however, concluded that reduction of advertising and solicitation expenses, the elimination of present bank difficulties in obtaining verification of automobile insurance renewals, the possibility of establishment of a “common pattern of operations” between loan officers and the single insurance agency, and certain “technical efficiencies in the data processing area” would constitute efficiency gains resulting in a public benefit.²¹ Significantly, in neither the First National nor the Southern order did the Board even mention the direct billing concept or the Administrative Law Judge’s reliance thereon.

After examination of the record, we find substantial evidence to support the Board’s conclusions that some of the types of efficiency gains it relied upon in the orders were reasonably likely to result. Although the briefs of the NAIA parties make abundantly clear that the evidentiary showings in favor of these conclusions were much weaker than they could have been, it is not our function, nor the Board’s, to require the strongest showing possible. Even without the professional studies, the detailed projections, and the unequivocal expert testimony that the NAIA parties deem essential, there is significant evidentiary support for the conclusions that there will be an avoidance of advertising and solicitation expenses, some elimination of duplicated information gathering and information sharing, and some efficiencies in data processing.

But the Board again has failed utterly in its responsibility to consider possible sources of efficiency loss and to arrive at a reasoned evaluation of net efficiency gains. It may well be that the Board could properly have concluded that a net gain would result or that, as First National suggests in its brief, the holding company affiliates would in the future adopt direct billing. But the fact remains that, despite the Administrative Law Judge’s explicit reliance upon the negative efficiency effects of non-use of direct billing, the Board failed to so much as mention the issue. Because of this failure, we cannot uphold the Board’s finding on this issue.

C. Increased Competition

Because insurance premium rates are set by insurance underwriters and subject to state regulation, competition on the basis of price among insurance agents is limited. Because of this, the Administrative Law Judge found that there was no “hard evidence” that there would be increased competition in the Georgia case. The Board, however, agreed with the Law Judge’s conclusion in the Alabama case that, although price competition would be limited, the applications would result in increased competition on the basis of service. In particular, the Board concluded that the greater convenience to consumers of “one-stop shopping” and the availability of bank financial expertise to insurance customers would force independent agents to increase their service efforts to compete with the holding company affiliates.

Here we find sufficient evidence to support the Board’s conclusions and no unexplained conflict with the Law Judge’s findings. To the evidence that there would be some pressure for lower prices, we must add the customer-attracting advantages of one-stop shopping and the financial expertise of loan officers, for which there is support in the record.²² While the evidence on this issue certainly can be described as vague, we find that in light of the realities of such applications, it is enough.

We find that this conclusion is bolstered by two factors not specifically relied upon by the Board. The first is that the entry by both Southern and First National into most of these lines of business will be de novo, rather than by purchase or merger with an existing competitor. See III App. 716; V App. 1230. It is quite clear that Congress “believe[d] that an activity commenced de novo will tend to have -procompetitive effects, and consequently should be viewed more favorably than the commencement of an activity through the acquisition of an existing concern”. Senate Report, 1970 U.S.Code Cong. & Adm.News at p. 5534. See Conference Report, 1970 U.S.Code Cong. & Adm.News at p. 5568; 12 U.S.C. § 1843(c)(8). Additionally, we take cognizance of the unusual competitive situation which presently obtains in the Georgia case. Citizens and Southern Holding Company, whose bank subsidiary has the largest share in the Atlanta banking market, already has authority to sell many of the types of insurance at issue here. See V App. 1349-50. It may reasonably be expected, then, that granting First National’s application will result in the introduction of further competition in the sub-area of bank-related insurance sales, and indeed will better enable First National to compete with Citizens and Southern in the provision of banking services as well.

D. Decreased or Unfair Competition

The primary claim of the NAIA parties with respect to this issue relates to the possibility that insurance sales will be “tied” to bank loans, thus reducing competition for reasons unrelated to efficiency. The claim is that loan officers may require prospective borrowers to insure through holding company affiliates as a condition for obtaining a loan (coercive tying) or that prospective borrowers may act on a belief that insuring through the holding company will be helpful or necessary in obtaining a loan, without being required to do so (voluntary tying).²³ The Administrative Law Judge noted that coercive tying was prohibited by the Holding Company Act Amendments of 1970, 12 U.S.C. § 1971 et seq.,24 but found that voluntary tying was “a distinct possibility”, particularly in times of tight money. The Board, however, found that there were no instances of voluntary tying described in the record, that the percentage of loan customers to which Tharpe & Brooks and a North Carolina bank holding company had sold insurance were not great, and that there were numerous other lenders, both bank and nonbank, to whom customers could look. It ruled that there was no reasonable likelihood of significant voluntary tying in either the Alabama or Georgia case and that the Law Judge’s requirement that specific statements regarding tying be incorporated in loan applications was not necessary.

The debate over voluntary tying²⁵ well illustrates the difficulty of the judgments entrusted to the Board under the Amendments. The Board was required to surmise the reasonable likelihood, in the future, of occurrences which are incapable of direct measurement even in the past. In its attempt to do so, the Board relied on two indirect indicia: past “penetration rates” and market structure. A penetration rate, which in this case is the percentage of borrowers from a particular lender who purchase insurance from the lender’s affiliate, is an indication of the upper limit of possible voluntary tying. Since borrowers may purchase the insurance from lenders for other reasons, such as greater convenience, a strong personal relationship with the lender, or better price or service, the actual rate of voluntary tying may be much lower than the penetration rate.

A First National officer testified that its subsidiary, Tharpe & Brooks, had experienced a penetration rate of 13.5 per cent with respect to mortgage lending. IV App. 830-31. And a North Carolina bank holding company also reported a low penetration rate by its insurance subsidiary operating a general insurance business. I App. 120. We agree with the NAIA parties that it would have been helpful to have further evidence on penetration rates, with information on rates with respect to particular types of loans, and with detailed consideration of the relevance of the North Carolina and Tharpe & Brooks rates to the applications at hand. But we find that this evidence is credible, significant support for the Board’s conclusion, particularly in view of the fact that the NAIA parties failed to present evidence or penetration rates contradicting the conclusions it suggested.

Second, the Board considered the structure of the respective lending markets, recognizing that the possibility of voluntary tying is inversely related to the availability of other sources of credit. In so doing, the Board responded to the Congressional belief that “the dangers of ‘voluntary’ tie-ins and reciprocity are basically structural”. Conference Report, 1970 U.S.Code Cong. & Adm.News at p. 5569. In the Alabama case, the evidence showed that there are six to twenty banking alternatives in each of the markets in which Southern subsidiaries operate, and that borrowers also have access to nonbank lenders such as finance companies, savings and loan associations, mortgage bankers, and credit unions. Ill App. 663-70. The Board in the Georgia case also relied upon the existence of other lenders as well as competition from over thirty banks in the Atlanta area. V App. 1233-34, 1327, 1330. We find these structural characteristics persuasive in support of the Board’s conclusions as to voluntary tying. Despite the fact that there are bank subsidiaries in the Alabama case which have leading shares in individual market areas, and that there may be particular types of commercial loans for which there are few alternative lenders, there was a reasonable basis for the conclusion that the total amount of possible voluntary tying was not of the magnitude Congress was concerned about. This, we believe, is the relevant inquiry, for the size of bank holding companies in 1970, as today, is such that many will have individual markets in which, for certain types of loans, they will be the only practical source of credit. To conclude that the existence of any individual instances of market hegemony necessarily requires a conclusion that voluntary tying is likely would require a finding of this adverse effect in a great number of applications. We find no evidence that Congress intended so particularized and stringent an inquiry, and hold that an overall estimation of the magnitude of likely tying is proper. The Board’s conclusion on this issue is supported by substantial evidence.²⁶

E. Undue Concentration of Resources

The NAIA parties challenge the Board’s conclusion that the size of the projected insurance agencies, the degree of concentration in the relevant banking markets, and the existence of alternative sources of loans do not result in an “undue concentration of resources”. It is not immediately clear to us the extent to which the factor of undue concentration differs from the factor of “decreased or unfair competition”. Our only real clue is the Conference Report’s statement that

[t]he dangers of undue concentration of resources include, but are not limited to, specific competitive effects, which are themselves relevant factors under the Act.

1970 U.S.Code Cong. & Adm.News at pp. 5567-68. We gather from this that Congress believed that concentration of economic resources in a single entity beyond a certain point was harmful regardless of the proven existence of any anticompetitive effects of such concentration.

The Board’s finding on this issue must be upheld. There can be no doubt either that the insurance businesses of the holding company affiliates will be relatively large or that the holding companies control a substantial amount of the banking resources in the relevant markets. But, as the NAIA itself emphasizes, the insurance agency business is one that is, in general, quite unconcentrated and one in which entry barriers are low. We note also that neither Southern nor First National subsidiaries are the largest banking entities in all but two of their respective markets. In view of the Board’s presumed expertise on this area and the fact that the determination of what is “undue” concentration of resources was primarily committed to the Board by Congress, we find that the Board must be upheld.

F. Conflicts of Interest

In the Alabama case, the Administrative Law Judge found that there was a reasonable possibility of conflicts of interest in the sale of surety bonds; the Law Judge believed that, since a line of credit is a prerequisite for such a bond, holding companies might grant or withhold credit on the basis of whether the customer had agreed to obtain the surety bond from their insurance affiliate. With respect to other types of insurance in the Alabama case, and with respect to all insurance involved in the Georgia case, the Law Judge found that there were no likely conflicts of interest other than those inherent in any sale of insurance where income is based on the amount of insurance sold. The Board found specifically that there was no credible evidence in the record that would support the finding as to surety bonds. The Board also found that the evidence did not indicate that the holding company banks would make undesirable loans to sell insurance and that regulatory supervision of banks would limit any misconduct of that type.

We are satisfied with the Board’s conclusions on these issues. Evidence in the record indicates that the insurance company affiliates will not be run on a pure profit maximization basis, and there is no evidence which would compel the opposite conclusion for which the NAIA parties contend. Contrary to the argument of the NAIA parties, the fact that a holding company’s interests as lender and as insurer do not totally converge does not require the conclusion that conflicts of interest will occur. The Board was entitled to credit the evidence that the insurance operations will be viewed as separate from and as an adjunct to the bank lending operations. While it would have been helpful for the Board to respond more fully to the Administrative Law Judge’s finding with respect to surety bonds, it is sufficient that the Board did recognize the Law Judge’s finding and indicate the source of its disagreement.

In summary, we find that the Board’s conclusions with respect to efficiency gains and greater convenience are not supported by substantial evidence, but that its other findings are. Accordingly, the Board properly found one type of public benefit, in the form of increased competition, and no adverse effects on the public interest. The Board’s overall finding, that public benefits outweigh adverse effects of granting the applications, can therefore be upheld.

V. Other Issues

A. Georgia and Alabama Law and the Specificity of Southern’s Proposal

The NAIA parties also contend that the Board gave insufficient consideration to a recently passed Georgia statute which, in essence, prohibits all lending institutions in sizable Georgia communities from selling insurance (other than credit life, accident, or health insurance). Ga.Code Ann. § 56-322 (Supp.1976). This statute, it is argued, indicates a public policy in the state against the very type of activity which the Board was asked to approve. We find the Board’s response convincing: .the statute by its own terms exempts lending institutions like First National and its larger competitor, Citizens & Southern Holding Company, from the prohibition through a grandfather clause. Id. § 56-322(d). Because of the clear exemption of First National from the statute, it cannot be said that either the letter or the policy of the Georgia statute warrants rejection of the First National application.

The NAIA parties also argue that the Board improperly failed to consider the effect of an Alabama law providing that only licensed insurance agents may take applications for insurance or place insurance for others. Ala.Code T. 28A, § 119(a) (Supp. 1974). Since Southern’s plan contemplated that only a single agent at its central insurance office would be licensed, there is a strong argument that the reference of bank customers to the agent by loan officers and the compilation of insurance information by the loan officers would be illegal. The Board’s opinion did not specifically mention the licensing statute, but apparently assumed that the information for insurance applications would be taken by licensed agents rather than by loan officers, a conclusion which seems to have little support from Southern’s own witnesses.

Our resolution of this issue is closely tied to that concerning another NAIA claim, that the Southern proposal was impermissibly vague and unformed. The reason for this connection is that the testimony of Southern’s witnesses made clear that the precise procedures by which insurance would be placed and applied for had not been clearly established by Southern. See II App. 232-35, 244; III App. 718-19. In this area as in others, we recognize that each side presents reasonable arguments. The NAIA parties are entitled to have a fairly specific proposal to challenge. Only then do they have a reasonable opportunity to protect their economic interests, which obviously are at stake, and only then can the public interests recognized by Congress in the 1970 Amendments receive realistic protection from the Board and the courts. Yet we can ask of Southern and other applicants only that which is reasonably possible. Where, as here, the holding company seeks to enter, de novo, into a type of activity with which it has no previous experience, it is wholly understandable that procedures for carrying on the activity may not have been codified in every detail. Indeed, in view of the difficult adjustments involved in entering a new business, particularly one which is heavily regulated, it may be the wisest policy to resist an early total commitment to one or another mode of operation.

Where the balance between these competing interests in specificity and flexibility should be struck is itself a question on which the Board must be considered to have some expertise. It knows far better than this court the time periods which applicants realistically may have to prepare their proposals, the amount of money which they can be expected to invest in them, and the degree of specificity which is inherently possible in each case. Indeed, there are no statutory or constitutional standards, other than those of due process, by which we can make these judgments. Accordingly, we believe that the matter is committed to the Board’s discretion and that we can overrule the Board on the issue of specificity only where there is a clear abuse of discretion or where there is a violation of due process. In view of the general specificity of Southern’s proposal and its lack of previous experience in insurance agency activities, we find neither in its failure to specify the details of proposed interaction between its banking and insurance operations.

With respect to the particular issue of licensing, we note that Southern acknowledged that its proposed procedures were less than fully fleshed out and repeatedly assured the Board that it would abide by all applicable licensing requirements. See III App. 718. Confident that the NAIA parties will scrutinize the Southern operation to ensure such compliance, we feel constrained to uphold the Board in its apparent assumption that Southern will operate not precisely as its witnesses suggested, but in a manner that will comply with Alabama law. To deny the Board the power to make this reasonable assumption would surely forfeit the benefits of the administrative process in terms of flexibility and common sense. We see no necessity for doing so.

B. Exclusion of Testimony

The NAIA parties complain of the Administrative Law Judge’s exclusion of portions of a 128 page written statement by Harry Houghton, an expert in the area of consumer finance. Part I of the statement dealt with the unique characteristics of banking and the market environment in which it is carried on, while Part II dealt with the nature of credit life and health insurance and the experience of lenders with the sale of such insurance. The Law Judge, who was upheld by the Board, concluded that the testimony was cumulative of the testimony of other witnesses (including Houghton’s oral testimony), was largely irrelevant to the issues in the proceedings, and primarily concerned credit life and health insurance, which the parties agreed was a permissible holding company activity. It was noted also that Mr. Houghton had submitted Part I to the Board in a previous proceeding.

So long as an administrative agency is not arbitrary, it has some discretion in determining whether to admit expert evidence. National Airlines v. CAB, 116 U.S.App.D.C. 114, 321 F.2d 380 (1963). This is particularly true with respect to evidence like that comprising Part I of Mr. Houghton’s statement, which is general and not directed to the specific controversy at hand. American Trucking Ass’ns, Inc. v. F. C. C., 126 U.S.App.D.C. 236, 377 F.2d 121 (1968). And it is permissible to reject that portion of the proffered evidence which has previously been accepted by the agency from the same party, Warren Co. v. NLRB, 353 F.2d 494 (1 Cir. 1965), cert. denied, 383 U.S. 958, 86 S.Ct. 1222, 16 L.Ed.2d 300 (1966). For all of these reasons, the exclusion of Part I of Mr. Houghton’s testimony was clearly proper.

Although the Board was correct in noting that the permissibility of the sale of credit life and health insurance was not a serious issue in these cases, the experience of lenders with that insurance was arguably relevant to the assessment of the likely public benefits and adverse effects of holding company activities with respect to other types of insurance. Particularly because of the dearth of relevant data upon which to base that assessment, it might have been preferable for the Administrative Law Judge to permit admission of that part of the statement. But we are unprepared to say that the action was so erroneous and prejudicial as to require reversal. As the NAIA parties themselves have strenuously asserted before us, credit life and health insurance differ from the types of insurance primarily involved in this case in several significant ways: they do not involve an insurance decision by the agent but are sold at a uniform low rate and, unlike property and liability insurance, are sold only because of an extension of credit. Because of these differences, which limit sharply the relevance of the experience with credit insurance to these cases, we cannot say that the Law Judge and the Board abused their discretion in excluding Part II of Mr. Houghton’s statement.

To summarize, we conclude that:

(1) we have jurisdiction to review § 225.-4(a)(9) of Regulation Y;

(2) § 225.4(a)(9) is procedurally valid;

(3) § 225.4(a)(9)(i) (except insofar as it authorizes the brokering of insurance for bank subsidiaries of bank holding companies, (ii)(c), and (iii) are invalid because of conflict with § 4(c)(8) of the Bank Holding Company Act;

(4) § 225.4(a)(9)(ii)(a) and (b) are valid;

(5) the Board’s orders in these cases are valid insofar as they authorize the sale of insurance for bank subsidiaries of the holding companies, property damage insurance, liability insurance, and other types of insurance authorized by the Board under § 225.-4(a)(9)(ii)(a) and (b);

(6) the Board’s conclusion that expected public benefits outweigh possible adverse effects is supported by substantial evidence;

(7) the Board did not err in its consideration of state law or in accepting Southern’s proposal despite its lack of specificity;

(8) the Board did not commit reversible error in excluding portions of the statement of Mr. Houghton.

AFFIRMED IN PART AND REVERSED IN PART.

1 . We note, however, that the legislative struggle continues even in the matters involved in this case. The “Competition in Banking Act of 1975”, (S. 2721) presently pending in Congress, would impose greater restrictions on bank holding company activities in non-banking areas than presently apply. See BNA Antitrust & Trade Reg. Rep. No. 754, at A-14 (Mar. 9, 1976).

2 . Having had experience with holding companies in other areas of commerce, Congress knew of the potential of such entities to multiply by many times the degree of concentration within an industry and to introduce unhealthy competitive practices in different but related industries. See, e. g., 15 U.S.C. § 79 et seq. (1970) (Public Utility Holding Company Act).

3 . The statutory provision referred to was § 4(c)(6) of the Act as passed in 1956. A 1966 amendment resulted in its renumbering.

4 . The Board excluded authorization for the sale of use and occupancy insurance, and business interruption and fidelity insurance because of lack of evidence. After a petition for review was filed in the Alabama case, the Board rescinded its authorization of the sale of mortgage guaranty insurance without Southern’s objection.

5 . As such, First National was not subject to the Bank Holding Company Act until the passage of the 1970 Amendments.

6 . The Board of Governors, in the Georgia case, stated that “in adopting the Insurance Regulation, [it had] found that the performance by bank holding companies of those activities authorized by such Regulation satisfied, in general, this public benefits test.” IV Consolidated Joint Appendix 736 (hereinafter “App.”). The NAIA parties argue that this indicates that the Board shifted the burden on the issue of net public benefits from the proposed activities from the applicant to the petitioners.

We agree that such a shift in the burden of proof by the Board would be error. See Conf. Rep. No. 91-1747, 91st Cong., 2d Sess., (1970) (Statement of the Managers on the Part of the House, hereinafter referred to as “Conference Report”), 1970 U.S.Code Cong. & Adm.News at p. 5570. The variables which Congress named as relevant to the public benefits determination are necessarily dependent upon local industry and market structures, upon the particular applications, and upon other factors which will vary widely in individual instances. While it may or may not be true that the Board makes a “general” public benefits determination in promulgating regulations on the “closely related” issue, see note 19 infra, any such determination should have no effect on the individualized inquiry which the Board must make in ruling on specific applications.

We find no indication, however, that the Board did shift the burden of proof to petitioners. Instead, the Board recognized that the public benefits determination required in connection with the specific applications was in addition to any general determination. IV App. 736. There is nothing in either of the Board’s orders that has been cited to us or that we have found which indicates that the Board did not act on its apparent belief that the individual public benefits determinations were to be additional to and separate from the generalized findings embodied in § 225.4(a)(9).

7 . This is true, for example, of the “convenience” insurance authorized by § 225.-4(a)(9)(ii)(c). Since the applications virtually track the language of this rather vague authorization, refusal to review the regulation would be tantamount to approval of the application on this point.

8 . It also may not be fair, in view of the fact that NAIA sought to raise many of these issues promptly upon the promulgation in September, 1972, of a regulation interpreting § 225.4(a)(9) of Regulation Y, 12 C.F.R. § 225.228. The Board in response strenuously and successfully argued that review of those regulations was at that time premature, since the issues could be raised in proceedings like this one! National Ass’n of Ins. Agents, Inc. v. Board of Governors of Federal Reserve System, 160 U.S.App.D.C. 144, 489 F.2d 1268 (1974).

We note also that considerations of concreteness and ripeness counsel that organizations like NAIA be permitted to wait until specific cases applying regulations arise before seeking review. In that way, courts like this one are presented with a more adequate record and more information as to the full ramifications of the regulations, and with argument by adversary parties other than the promulgating agency. All of these factors aid the court in reaching an informed decision. See K. Davis, Administrative Law Treatise § 21.01 (1958).

9 . The Board ruled on 27 applications to engage in insurance activities during this period.

10 . See, e.g., General Contract Corp., 1959 Fed. Res.Bull. 260, Transamerica Corp., 1957 Fed. Res.Bull. 1014 (involving insurance underwriting rather than agency).

11 . Sen.Rep. No. 91-1084, 91st Cong., 2d Sess. 12 (1970), 1970 U.S.Code Cong. & Adm.News at p. 5530 (hereinafter referred to as “Senate Report”), provides:

First, the language of the act will be amended to allow the Federal Reserve Board greater flexibility in determining what nonbanking activities and acquisitions are to be permitted bank holding companies.

The House Report, H.R.Rep. No. 91-387, 91st Cong., 2d Sess. 14 (1970) (hereinafter referred to as “House Report”), similarly states:

The committee amendment gives the Federal Reserve Board added flexibility in administering the Act in terms Chairman Martin himself has suggested.

12 . This language was inserted at the request of the then-Chairman of the Federal Reserve Board, William McChesney Martin. See Conference Report, 1970 U.S.Code Cong. & Adm. News p. 5566; Senate Report, 1970 U.S.Code Cong. & Adm.News p. 5531.

13 . Those members of the Conference Committee signing the Conference Report were Rep. Wright Patman (D-Tex.), Chairman of the House Banking and Currency Committee, Rep. William A. Barrett (D-Pa.), Rep. Leonor K. Sullivan (D-Mo.), and Rep. Henry Reuss (D-Wis.). The Congressional practice in 1970 was that only House members would write a “statement of managers” accompanying Conference Reports. See 116 Cong.Rec. 42422 (1970).

14 . 116 Cong.Rec. 41952-53 (remarks of Rep. Widnall); id. 42422, 42424 (remarks of Sen. Sparkman); id. 42431 (remarks of Sen. Bennett).

15 . The latter change in particular was sought by the Board to make it clear that nonbanking activities need not relate to the particular banking business of the applicant but to banking in general, thus permitting activities in areas geographically removed from the applicant’s bank operations. See Conference Report, 1970 U.S. Code Cong. & Adm.News pp. 5566-67. There is persuasive evidence that the insertion of the “functionally related” language was intended only to accomplish this, and not to effect a general broadening of permissible activities. See 116 Cong.Rec. 42434 (1970) (letter from Arthur F. Burns, Chairman of the Federal Reserve Board).

16 . E. g., Bank Shares, Inc., 45 Fed.Res.Bull. 954 (1959); First Bank Stock Corp., 45 Fed.Res. Bull. 917 (1959).

17 . Petitioners have not specifically attacked this portion of § 225.4(a)(9), but in view of their general contention that insurance agency activities other than the sale of credit life, and credit health and accident insurance are not “closely related” to banking, Brief for Petitioners at 45, 49, Alabama Ass’n of Ins. Agents, Inc. v. Board of Governors; Brief for Petitioners at 46, 51, Georgia Ass’n of Ind. Ins. Agents v. Board of Governors, it is our duty to determine its validity.

18 . T. I. M. E., Inc. v. United States, 359 U.S. 464, 79 S.Ct. 904, 3 L.Ed.2d 952 (1959); Shapiro v. United States, 335 U.S. 1, 68 S.Ct. 1375, 92 L.Ed. 1787 (1948); United States v. Douglas Aircraft Co., 510 F.2d 1387 (Ct.Cust. & Pat.App.1975); Kansas City v. Federal Pac. Elec. Co., 310 F.2d 271 (8 Cir. 1962), cert. denied, 371 U.S. 912, 83 S.Ct. 256, 9 L.Ed.2d 171 (1963).

19 . As noted earlier, the Board in its order in the Georgia case evidenced a belief that its determination in § 225.4(a)(9) that certain insurance activities are closely related to banking also served as a determination that those activities “in general” satisfy the public benefits test. Without ruling on the correctness or incorrectness of this statement, we note that Chairman Burns of the Federal Reserve Board indicated in a letter to the Conference Committee his belief that the

language of the statute should make clear that there are two tests to be satisfied in deciding whether nonbank subsidiaries of bank holding companies should be allowed to engage in a particular activity: (1) whether the activity is closely related to banking, and (2) whether it is a proper incident to banking, and that in applying the second test, the Board is to weigh the possible benefits to the public against the risks of adverse consequences. We believe that the language proposed by the House conferees might be misconstrued to apply three tests: (1) whether the activity is closely related, (2) whether it is a proper incident to banking, and (3) whether the benefits outweigh the risks.

116 Cong.Rec. 42434 (1970). The Chairman, whose suggestions on this matter were heeded by the Committee, thus seems to have regarded the “closely related” and “proper incident” determinations as wholly separate.

20 . Southern, on the other hand, has referred customers to a number of insurance agencies upon request. II App. 229.

21 . II App. 193-94; IV App. 737. We note initially that there is some dispute over the precise character of the efficiency gains referred to in the statute. The NAIA parties apparently contend that gains in efficiency are not relevant for the purpose of the statute unless the applicant can demonstrate that they will be passed along to the consumer. We find no basis in the statute for this argument; the statute requires consideration of “gains in efficiency” and apparently assumes that such gains will constitute “public benefits”.

To be sure, improved service or lower prices to the consumer are one type of public benefit, but it is also true that a public benefit in a broad sense is conferred whenever a service is performed in a new way which requires less of society’s limited human and economic resources. The result may be lower prices to the consumer of the service, or it may be greater dividends to the body of stockholders of the service provider, better salaries or benefits to the provider’s employees, or the availability of greater capital for expansion of the business. In the latter instances as well as the former we believe that efficiencies reasonably can be said to result in a public benefit: the freeing of scarce resources for beneficial economic use.

22 . See, e. g., II App. 249-51, 283-84, 307-08, 311, 313; III App. 562, 592-95; IV App. 768-70, 811, 836-37, 838-39, 842^4, 863-64, 869, 880-81; V App. 1225, 1227-32, 1268, 1274-77. Even though the finding of greater convenience could not be upheld, we must note the evidence in support of that conclusion as one factor tending toward a conclusion that service competition will result. This is so because the likelihood of increased service competition depends upon consumer perceptions of convenience and the response of independent agents to this perception, rather than the actual results concerning increased convenience.

23 . As the NAIA parties note, there is a spectrum of possible activities between overt coercive tying and purely voluntary tying. Bank loan officers may pointedly question about the prospective borrower’s intended insurance agent at a critical point, strongly recommend the bank-affiliated agent prior to making the loan decision, or use other effective but indirect pressuring tactics.

24 . 12 C.F.R. § 225.4(c) extends the prohibition against coercive tying from banks to other bank holding company subsidiaries approved under § 4(c)(8) of the Bank Holding Company Act.

25 . We find that the Board’s conclusion as to coercive tying is amply supported by the evidence. Not only is coercive tying specifically prohibited by law (and both applicants have pledged to observe the law), but the Board’s Regulation Z, 12 C.F.R. § 226.4(a)(5), (6), requires disclosure to lenders that the customer may choose the person through which any insurance is to be obtained. More importantly, the NAIA parties were able to produce no evidence of coercive tying on the part of either applicant.

26 . The NAIA parties also argue that a finding of likely voluntary tying was mandated by testimony of a Southern officer that, in commercial lending situations only, the borrower’s insurance arrangements are scrutinized prior to the closing of the loan. The purpose of such inquiry is to make sure that the proper coverage has been arranged, in order to prepare major loan proposals for comprehensive scrutiny of a loan committee. The officer who testified on this matter added that the necessary insurance often has been obtained by the time the proposal has been brought to the bank. In light of this fact, we cannot say that the number of instances in which the possibility of voluntary tying is increased is sufficiently large to require rejection of the Board’s conclusion.