On January 13th, the Small Business Administration (SBA) issued proposed regulations to implement the anti-bundling provisions of P.L. 105-135, the Small Business Reauthorization Act of 1997. 64 Fed. Reg. 2153 (Jan. 13, 1999). These proposed regulations follow closely on the heels of two high-profile General Accounting Office (GAO) decisions on the subject of bundling, and a less well-publicized decision from the United States Court of Federal Claims, and they highlight some of the difficulties Federal agencies face as they take advantage of their discretion to engage in commercial buying practices. Bundled procurements are generally more efficient for the procuring agency; they often appear to save money; and in some cases they are clearly necessary in order to support specific technical, quality, or design requirements. But they achieve these goals at a price. Bundled procurements reduce competition, and in a system where competition is the guiding principle, balancing of the benefits of bundling against its disadvantages is tricky.
Just what is a bundled procurement? And when do its benefits outweigh its disadvantages? GAO and the courts have struggled with these questions, and SBA is now struggling with them in its proposed regulations. Whatever the impact of the final regulations, it is clear that bundling is an issue for both large and small businesses, and that it is not going away in the near future.
What Is Bundling?
This is the $64,000 question. SBA proposes to define the term this way:
The term “bundled requirement or bundling” refers to the consolidation of two or more procurement requirements for goods or services previously provided or performed under separate smaller contracts into a solicitation of offers for a single contract that is likely to be unsuitable for award to a small business concern due to-
(A) The diversity, size, or specialized nature of the elements of the performance specified;
(B) The aggregate dollar value of the anticipated award;
(C) The geographical dispersion of the contract performance sites; or
(D) Any combination of the factors described in paragraphs . . .(A), (B), and (C).
13 CFR § 125.2(d)(1)(i) (proposed), 64 Fed. Reg. 2156 (1999).
This is a good working definition in the context of small business policies, but it ignores the fact that bundling routinely occurs in procurements in which large businesses are the only likely participants. For example, bundling can occur in large construction projects, or indefinite delivery/indefinite quantity (IDIQ) contracts, or task order contracts, where the principal competitors are large businesses. In this context, the “policy” concern is simply competition (or the lack of it), without any small business overlay.
Regardless of how the term is defined, bundling’s effect is to group a variety of segregable requirements into a single solicitation, so that all bidders/offerors must be able to satisfy all of the requirements. Offerors that can provide only some are excluded from the competition.
When Is Bundling Justified - And When is It Improper?
The backdrop for this discussion is the Competition in Contracting Act, which requires full and open competition and states clearly that solicitations may contain restrictive provisions and conditions only to the extent necessary to satisfy the needs of the agency, or as authorized by law. 10 U.S.C. § 2305(a)(1)(B)(ii); 41 U.S.C. § 253a(a)(2)(B). Since bundling restricts competition, when is it permissible?
There are three places to look for an answer. First, we can look at bid protest decisions, most of which have come from GAO. Over the years, GAO has evaluated bundled procurements in a variety of contexts, and with a few important exceptions, it has come down squarely on the side of competition and against bundling. For comparison, we can also look at the Federal Acquisition Regulations (FAR) implementing statutory requirements on multiple awards under IDIQ contracts. The FAR lists the limited circumstances in which an agency may be able to justify issuing a single, bundled award for an IDIQ contract, rather than issuing multiple awards. And finally, we can now look at SBA’s proposed regulations, and the statutory language on which they are based.
GAO’s Answer: Competition Prevails Over Bundling Except In A Few Limited Circumstances
GAO has taken a dim view of bundling, a fact that is evident from a review of bid protest decisions on the subject. Each case is fact-specific, but most of these protests are sustained because of the deference accorded to statutory competition requirements. Moreover, GAO has been sensitive to the adverse effects of bundling on competition, not only in connection with the procurements that are the subject of its protests, but also in connection with potential future procurements. Two decisions involving supply contracts are illustrative.
Allfast Fastening Systems, Inc., B-251315, Mar. 25, 1993, 93-1 CPD ¶ 266, involved a solicitation that bundled 97 different types of rivets into 4 lots. The protester could not provide 9 of the 97 types of rivets in question, and this precluded it from submitting a proposal for any of the 4 lots. The protester asserted, and GAO clearly believed, that its presence as an additional competitor over the years had resulted in significant cost savings to the Government, and it suggested as a remedy in this situation, that merely taking 4 of the rivet types out of 2 of the lots would enable it to compete for those 2 lots, thereby increasing competition and reducing the price. In GAO’s view:
In addition to the savings . . ., increasing competition here would also lead to the continued potential for long-term savings by maintaining a viable base of competition for these rivets. . . . DISC has failed to respond with any rationale for why the agency cannot make this or some similar adjustment to increase the limited competition available here. . . .
Id., 93-1 CPD, at 6. Thus GAO’s analysis took into account not only competition on the instant procurement, but the potential for future competition as well.
This concern for immediate and future competition was also apparent in Pacific Sky Supply, Inc., B-228049, Nov. 23, 1987, 87-2 CPD ¶ 504. There, GAO sustained a protest against bundling of 294 spare parts for T-56 aircraft engines for a five-year contract period, which effectively created a sole-source award since only one contractor was able to supply all 294 items. The agency attempted to justify the bundled requirements on grounds of administrative convenience-fewer contracts to administer-and also strongly argued its own commitment to “aggressively pursue” alternate sources for various line items during the life of the contract. GAO accepted the protester’s argument that the administrative burden on the agency was too insubstantial to justify bundling, and GAO opined that a five-year contract period was simply too long to lock in an anti-competitive situation, notwithstanding the agency’s commitment to seeking out additional sources for future procurements.
Bundling, then, is strongly discouraged, carefully scrutinized, and seldom upheld in a protest. There are, however, a few situations in which bundling has been adequately justified, at least in GAO’s view. Most of those situations fit into one of three categories:
(i) Design integrity/interoperability: the requirements to be bundled all relate to a single integrated system, where design integrity and interoperability are critical.
If the requirements to be bundled all relate to a single integrated system, where design integrity and interoperability are critical, GAO has generally allowed the bundled procurement to go forward. For example, in Titan Dynamics Simulations, Inc., B-257559, Oct. 13, 1994, 94-2 CPD ¶ 139, bundling of requirements for pyrotechnic simulators with the balance of the Navy’s requirements for the Multiple Integrated Laser Engagement System 2000 was allowed, despite the protest of a small business that wanted to bid on the simulators by themselves. GAO accepted the agency’s argument that unbundling the pyrotechnic subsystems from the rest of the system design effort ultimately could affect the integrity of the total system. The protest was denied, and the solicitation was permitted to continue on a bundled basis, even though the protester was excluded from the competition.
Similarly, in Building Systems Contractors, Inc., B-266180, B-266184, Jan. 23, 1996, 96-1 CPD ¶ 18, GAO denied a protest which asserted that the Air Force had improperly bundled a computerized energy management control system (EMCS) with the HVAC system and installation services to which it applied. The Air Force was able to identify a series of specific benefits it would gain from use of the brand name EMCS, including design and operations compatibility with its existing system. The protest was denied. See also Magnavox Electronic Systems Co., B-258037, B-258037.2, Dec. 8, 1994, 94-2 CPD ¶ 227; Resource Consultants, Inc., B-255053, Feb. 1, 1994, 94-1 CPD ¶ 59; Space Vector Corp., B-253295, B-253295.2, Nov. 8, 1993, 93-2 CPD ¶ 273; Electro-Methods, Inc., B-239141, B-239141.2, Nov. 5, 1990, 90-2 CPD ¶ 363.
(ii) Overwhelming administrative burden: the agency’s administrative burden will be truly overwhelming if the work is not bundled.
Although agencies often defend solicitations that involve bundled requirements on grounds of administrative efficiency, they seldom win. See, e.g., Better Service, B-265751, B-265751.2, Jan. 18, 1996, 96-1 CPD ¶ 90; National Customer Engineering, B-251135, Mar. 11, 1993, 93-1 CPD ¶ 225. GAO’s threshold for administrative convenience is low, and its tolerance for inconvenience to the agency is high, at least when balanced against the statutory requirement for competition:
when concerns of economy or efficiency are being weighed against maximizing competition, CICA and the implementing regulations require that the scales be tipped in favor of maximizing competition.
National Customer Engineering, 93-1 CPD, at 6.
If the burden to be avoided rises to the level of financial absurdity, however, GAO has sometimes denied the protest and allowed the bundled procurement to go forward, particularly where the agency was under severe budgetary and personnel constraints. Advanced Elevator Services, Inc., B-272340, B-272340.2, Sept. 26, 1996, 96-2 CPD ¶ 125, is a case in point. There, GSA issued a negotiated solicitation that contemplated award of up to 5 contracts for elevator maintenance services at 178 federal buildings in 3 GSA regions that spanned 15 states. The protester objected that bundling the elevator service requirements into 5 multiple-region contracts was unduly restrictive of competition, since those requirements were larger than any small business firm could provide.
In response, GSA elaborated on its administrative burden: recent budgetary problems, which had resulted in significant personnel losses, left the agency without sufficient personnel and resources to continue to award and manage more than 100 separate contracts, as it had done in the past. GAO found this explanation reasonable and well supported, and it denied the protest. See also The Sequoia Group, Inc., B-252016, May 24, 1993, 93-1 CPD ¶ 405.
The facts in this case were egregious, and the budgetary/personnel problems that GSA was attempting to solve through a bundled approach to its procurement created a situation more akin to financial impossibility than mere “administrative burden.” But the case does confirm that, when an agency demonstrates that the administrative burden of pursuing multiple unbundled procurements would be truly overwhelming, GAO has supported the resulting restrictions on competition.
(iii) National security: the agency can provide a well-documented justification based on national security or military readiness considerations.
Two cases decided last fall highlight bundling in the context of military readiness concerns. National Airmotive Corp., B-280194, Sept. 4, 1998, 98-2 CPD ¶ 60, and Pemco Aeroplex, Inc., B-280397, Sept. 25, 1998, 98-2 CPD ¶ 79, decided within 3 weeks of each other, both involved bundling of depot maintenance and repair services for the Air Force. The National Airmotive solicitation sought depot-level maintenance and repair of T-56, TF-39, and F-100 aircraft engines. The Pemco Aeroplex solicitation sought depot maintenance for the KC-135 aircraft, inspections and painting of the A-10 aircraft, and overhaul and repair requirements for hydraulic components, electrical accessories, and flight instruments. In both cases, the Air Force defended its decision to bundle the requirements on grounds of potential adverse impact to readiness if the requirements were not bundled, and its enhanced ability to handle surge demands if the requirements were bundled. GAO sustained the protest in Pemco Aeroplex, but denied it in National Airmotive.
A distinction supporting GAO’s opposite conclusions in these cases rests on the quality of the evidence used by the Air Force to prove its readiness concerns. In National Airmotive, the agency pointed to specific data on “war readiness engines,” which showed a degraded level of readiness at the time of award, as well as documented attrition rates at the base in question that resulted from Base Realignment and Closure Commission recommendations, and which translated directly into decreases in productivity as work was transitioned. Based on this information, GAO concluded that bundling the engine maintenance requirements for three different engines, which involved a “substantial number of common processes” as well as a substantial amount of common equipment used to perform the workloads for the three engines, was a reasonable approach despite the reduction in overall competition. “[W]hen legitimate, properly documented, national security concerns are weighed against interests in maximizing competition, the scales must tip towards national security.” National Airmotive, 98-2 CPD, at 10.
By contrast, the Air Force’s similar arguments in Pemco Aeroplex were unpersuasive. In National Airmotive the Air Force provided specific data on its degraded level of readiness; here there was no such data. In National Airmotive, the Air Force demonstrated that there were significant commonalities among the various workloads that would benefit from performance by a single contractor; here the commonality was substantially less. Moreover, the Air Force provided no persuasive evidence to support its surge argument, and its claims of likely losses in productivity from multiple transitions did not address why careful management could not minimize or avoid that problem. GAO explained that, “[i]n sum, while our office will show deference to agency claims that requirements of military readiness support a combination of workload requirements . . ., such claims must be properly documented and reasonably related to the workload requirements.” Pemco Aeroplex, 98-2 CPD, at 13. The evidence on these points in Pemco Aeroplex was, in GAO’s view, insufficient.
Despite the different results, these two cases stand for the proposition that GAO is willing to consider military readiness as a justification for a bundled procurement, although it will require concrete evidence to prove that any readiness concerns are both documentable and documented.
Aside from cases that fall into one of the three broad categories described above-design integrity, overwhelming administrative burden, or national security-GAO has rarely denied protests against bundling. Competition has been King. Thus the first “answer” to the question of “when is bundling permitted?” is that GAO has permitted bundled procurements to go forward only when one of these three circumstances exists. The GAO decisions, however, do not provide the only source of guidance on this question.
IDIQ Contracts Are Governed By Slightly Different Considerations
Historically, the “bundling” issue has arisen in garden-variety supply, service, or construction contracts, but more recently the concept has spilled over into IDIQ and task order contracts. In fact, amendments to the Competition in Contracting Act (CICA) that were passed in 1994 as part of the Federal Acquisition Streamlining Act created: (i) a scheme for “multiple awards” under IDIQ procurements, which is effectively a statutory preference in favor of unbundling, and (ii) a requirement for unbundling on task order contracts for advisory and assistance services where the amount is expected to exceed $10 million over 3 years. 10 U.S.C. § 2304a; 41 U.S.C. § 253h; see also Nations, Inc., B-272455, Nov. 6, 1996, 96-2 CPD ¶ 170. Legislative history clearly focuses on the benefits of constant, head-to-head competition among multiple awardees as tasks or delivery orders are identified and released. See S. Rep. No. 103-258, at 15-16 (1994), reprinted in 1994 U.S. Code Cong. & Admin. News 2561, 2576.
These statutory provisions have been implemented through the FAR at 16.500 et seq. FAR 16.504(c)(1) establishes the general preference for multiple awards on IDIQ contracts, but identifies six situations in which multiple awards “should not” be made. The six situations borrow heavily from the case law on bundling, and (except for the last two, which involve the simplified acquisition threshold and a catchall “in the best interests of the Government” provision that would presumably cover national security) are generally recognizable as restatements of the arguments put forth by various agencies over the years in support of their bundling determinations:
Multiple awards should not be made if the contracting officer determines that-
(i) Only one contractor is capable of providing performance at the level of quality required because the supplies or services are unique or highly specialized;
(ii) Based on the contracting officer’s knowledge of the market, more favorable terms and conditions, including pricing, will be provided if a single award is made;
(iii) The cost of administration of multiple contracts may outweigh any potential benefits from making multiple awards;
(iv) The tasks likely to be ordered are so integrally related that only a single contractor can reasonably perform the work;
(v) The total estimated value of the contract is less than the simplified acquisition threshold; or
(vi) Multiple awards would not be in the best interests of the Government.
Except for advisory and assistance task order contracts expected to exceed $10 million (where multiple awards are required), these six criteria provide some guidance on when bundling is justified-or not justified-at least for IDIQ contracts. With IDIQ contracts, bundling into a single award is likely to be permitted when an agency can demonstrate that only one contractor can provide the requirement, or if administrative costs of multiple awards are overwhelming, or if design integrity or interoperability are critical, or if “more favorable terms and conditions” will prevail with a single award.
The analysis does not stop there, however, since GAO has grafted a “void for vagueness” standard onto IDIQ solicitations. In fact, in connection with a protest that it dismissed as untimely last year, GAO took the extraordinary step of writing a letter to the Air Force and the Army, chastising those agencies for violation of CICA and FAR 16.504 even though it did not address the protest on the merits. Letters to the Air Force and Army Concerning Valenzuela Engineering, Inc., B-277979, Dec. 9, 1997, Jan. 26, 1998, 98-1 CPD ¶ 51. GAO reviewed the statement of work in this IDIQ solicitation and concluded that it was so broad that it did not “reasonably describe the scope of services needed,” thus failing to provide potential offerors notice of the work that would be within the scope of the resulting contract. GAO took pains to point out that:
inclusion of broad categories of work in one statement of work constitutes a form of bundling, since different kinds of work (or tasks in different geographical or technical areas) are combined into one procurement, and an overly broad statement of work can unjustifiably diminish competition, just as bundling does, by deterring businesses, particularly small businesses, from competing for a contract, notwithstanding their ability to perform some of the work at issue.
Id., 98-1 CPD, at 2 (emphasis added). Thus a solicitation with a broad, vague, undifferentiated statement of work can constitute a prohibited form of bundling, at least in connection with IDIQ contracts. GAO has not addressed whether this would be true even if the agency could make persuasive arguments in favor of a single, bundled award on grounds of quality, design integrity, or national security.
SBA’s Proposed Regulation Adds A Few New Wrinkles
Amendments to the Small Business Act passed in 1997 (P.L. 105-135) are driving the SBA’s proposed regulation. Section 413 of the Small Business Reauthorization Act of 1997 (now codified at 15 U.S.C. § 644) directs that:
[C]onsolidation of the requirements [in a solicitation] may be determined as being necessary and justified if, as compared to the benefits that would be derived from contracting to meet those requirements if not consolidated, the Federal Government would derive from the consolidation measurably substantial benefits, including any combination of benefits that, in combination, are measurably substantial. Benefits described in the preceding sentence may include the following:
(i) Cost savings.
(ii) Quality improvements.
(iii) Reduction in acquisition cycle times.
(iv) Better terms and conditions.
(v) Any other benefits.
15 U.S.C. § 644(e)(2)(B) (emphasis added). SBA’s proposed regulation is slightly more expansive; it states that “measurably substantial benefits include any one, or more, of the following in any combination, or in the aggregate:”
(1) Cost savings and/or price reduction;
(2) Quality improvements that will save time or improve or enhance performance or efficiency;
(3) Reduction in acquisition cycle times;
(4) Better terms and conditions; or
(5) Any other quantifiably substantial benefits
13 CFR 125.2(d)(3)(iii)(A) (proposed), 64 Fed. Reg. 2157 (1999).
Setting aside for the moment the issue of “measurably substantial benefits,” it is clear that this list of criteria for evaluating when bundling is justified differs in some respects from the trends identified in GAO decisions. For example:
GAO’s occasional sensitivity to “overwhelming administrative burden” does not show up in this listing, although a subsequent section of the regulation states that if “administrative or personnel cost savings are expected to be substantial”, bundling may be justified. 13 CFR 125.2(d)(3)(iii)(B) (proposed), 64 Fed. Reg. 2157 (1999).
GAO’s general support for design integrity and interoperability considerations may be reflected in subpart (2) of this proposed regulation (“quality improvements that will .. enhance performance or efficiency”), but this is subject to argument.
“National security considerations” are not specifically identified, although they could be included in the catchall of subpart (5), for “any other quantifiably substantial benefits”.
While the SBA regulation identifies “cost savings” as its first consideration, GAO has never regarded projected cost savings by itself as a sufficient justification for bundling of requirements into a single solicitation. To the contrary, GAO’s operating assumption has been that competition inevitably lowers prices, while bundling is likely to raise them. See, e.g., Allfast Fastening Systems, supra.
Perhaps more significantly, the emphasis in this proposed regulation departs from what has been the emphasis in the GAO decisions. GAO’s analysis, driven by CICA, has been grounded in competition; only when an extraordinary set of facts demands it has GAO allowed a bundled procurement to proceed, since bundling restricts competition. The approach taken by Congress in the statute and by SBA in the proposed regulations, however, shows only passing interest in competition as a principle. Instead, the primary consideration-the principal focus-is on process, i.e., maximizing small business access to federal procurements. Toward this end, the proposed regulations impose a strict quantitative analysis requirement. The exercise of judgment and discretion by an agency-to which GAO typically defers-is far more tightly circumscribed than it has been in the typical GAO decisions. Under the proposed regulation, the effect of bundling must be quantified and its justification must involve “measurably substantial benefits.”
The notion of quantitative analysis is nevertheless foreshadowed in some of the more recent GAO decisions, and in a recent decision from the United States Court of Federal Claims on an IDIQ contract. In the two “national security” decisions discussed above, Pemco Aeroplex and National Airmotive, GAO’s analysis showed a strong preference for quantitative evaluation of benefits and risks by the agency-and where that kind of analysis was lacking, as in Pemco Aeroplex, the protest was sustained, and the agency was directed to unbundle the procurement. Perhaps GAO’s decision in these two cases was informed by the statutory language in P.L. 105-135 (“measurably substantial benefits”), which had been passed almost 9 months earlier, even though the legislation did not apply to the procurements at issue there.
The United States Court of Federal Claims traveled down this same path in Winstar Communications, Inc. v. United States, 41 Fed.Cl. 748 (1998). The agency there had issued an IDIQ solicitation, stating its intention to make a single award. No written justification for this strategy was prepared, and the court found that the agency had simply failed to do any cost-benefit analysis as required by FAR 16.504(c):
The CO did not address these benefits in analysis. In addition, he stated during his deposition that he was not aware of an attempt by GSA to determine the price or other benefits which might accrue if multiple awards were made.
41 Fed.Cl., at 759. It is not surprising that the Court sustained this protest due to the lack of analysis by the agency. What is interesting, however, is just how closely the Court followed the dictates of FAR 16.504(c), and how little deference it gave to the agency’s arguments that documenting such an analysis would not affect the outcome.
Seen in the light of these decisions, the SBA’s proposed rules on bundling may not be as great a departure from precedent as they appear. However, at least in connection with procurements where small business participation is a possibility, they will add another layer to the analysis of when “bundling” may be justified, and when it is improper.
— Pat Wittie
Copyright © 1999 Kilcullen, Wilson & Kilcullen. All rights reserved.