Soothsaying Oracle: What GAO’s Decision Concerning DoD Prototype OTAs Means for DoD and Beyond

Federal ProcurementOther Transaction Agreements (OTAs):  The term is familiar among government contractors, yet commonly misunderstood.  Recently, the Government Accountability Office (GAO), in Oracle America, Inc., B-416061, May 31, 2018, 2018 CPD ¶ ___, defined OTAs as “legally-binding instruments, other than contracts, grants, or cooperative agreements, that generally are not subject to federal laws and regulations applicable to procurement contracts.”  Breaking this down, according to GAO, OTAs are:  (1) legally enforceable; (2) not contracts (or grants or cooperative agreements); and (3) not subject to procurement laws and regulations.  Importantly, however, OTAs in fact are contracts, in that they are legally enforceable agreements between two parties, under which the Court of Federal Claims has determined one party may sue the other for breach of contract.  But OTAs are not procurement contracts, or agreements under which the Government acquires goods or services for its direct benefit.  By statute, if the Government purchases goods or services for its benefit, the agreement effectuating that transaction must be a procurement contract, and as such must be subject to all applicable federal procurement laws and regulations.  Or at least this was the general rule, until GAO departed from it in MorphoTrust USA, LLC, B-412711, May 16, 2016, 2016 CPD ¶ 133, a departure reaffirmed in Oracle.

Much of GAO’s discussion in the Oracle decision applies to DoD’s unique authority to enter into OTAs to acquire prototypes.  DoD’s prototype OTAs are different from other OTAs, primarily because they are “acquisition instruments,” through which DoD acquires goods and services for its direct benefit.  These would classically be considered procurement contracts, but for Congress’ direct authorization to the contrary.  No other federal agency has been authorized to use OTAs for this purpose (putting aside, for the moment, GAO’s departure in MorphoTrust).  Instead, other agencies’ OTA authorities allow them to promote commercial development, in particular with non-traditional government contractors (e.g., NASA’s Space Act Agreements to develop commercial spaceflight capabilities) or to enter into arrangements that otherwise are not feasible with a typical procurement contract (e.g., TSA agreements with local law enforcement for airport security).

The unique nature of the prototype OTA as an acquisition instrument goes hand-in-hand with DoD’s additional authority to provide for follow-on production work in the prototype OTA itself.  It is this specific authority to secure follow-on production work through a prototype OTA that was at issue in Oracle.  This does not mean the decision has limited impact.  DoD is one of the most active agencies awarding OTAs, and GAO’s discussion of DoD’s authority provides critical insight about how GAO might examine the OTA authority of other agencies in the future.  Below, we examine the practical implications of GAO’s commentary on DoD prototype OTAs and follow-on production awards, then consider whether GAO’s standard in Oracle gives other agencies a green light to enter into acquisition-like OTAs.

Oracle: GAO’s First Foray into DoD’s Prototype OTA Authority

DoD’s prototype OTA authority enables it to acquire the exact prototypes it wants without the regulatory burdens of a procurement contract.  As such, prototype OTAs have become a favorite tool of DoD as it seeks to engage commercial firms to develop new defense technologies.  The Defense Innovation Unit (Experimental) (DIUx), DoD’s outpost in Silicon Valley, leads the way with this mission.  One innovative acquisition method championed by DIUx is its Commercial Solutions Openings (CSOs), through which DoD entities post solicitations for new technologies and DIUx facilitates the award of a prototype OTA, often with the goal of eventually awarding a full production contract or production OTA.

In Oracle, GAO examined whether the Army properly awarded, without competition, a $950 million production OTA for cloud migration and operation services.  The services had been tested in a prototype OTA awarded through one of DIUx’s CSOs just nine months earlier.  GAO concluded the Army’s award was not proper for two reasons.  First, the Army had failed to “provide for” the follow-on award in the initial OTA.  Second, prototyping under the initial OTA was not yet complete.

Providing for the follow-on award in the prototype OTA and completing the prototype prior to production are both statutory prerequisites to DoD’s exercise of its follow-on production authority.  Specifically, 10 U.S.C. § 2371b, paragraph (f)(1) allows DoD to enter into prototype OTAs that “provide for the award of a follow-on production contract or transaction to the parties in the transaction,” while paragraph (f)(2) requires “competitive procedures” to be used in the award of the initial prototype OTA.  Paragraph (f)(2) also requires the prototype project to be “successfully completed” before DoD can make a follow-on award.

GAO, without getting too far into the extent of the competitive procedures required in an initial OTA award (but noting the CSO process met it), determined that the prototype OTA at issue in Oracle did not “provide for” follow-on production work because the language of the OTA itself did not contain a provision contemplating a follow-on award.  It was not enough, GAO concluded, that DIUx’s CSO referenced the possibility of a follow-on; the prototype OTA itself must contain the magic language.

In determining that prototyping had not finished, GAO pointed to modifications the Army had issued during performance that added work to the OTA and noted that work under these modifications still was ongoing.  GAO determined that completion must be measured against the full and current scope of work described in the prototype OTA, including all modifications.  Even though all prototyping under the original scope of work had finished, GAO concluded that the OTA awardee was not yet eligible for a noncompetitive follow-on.

In addition to these two successful protest grounds, GAO also considered Oracle’s argument that the initial OTA was not really for a “prototype.”  GAO found this less persuasive.  It rejected Oracle’s dictionary definition, which would have excluded the OTA’s subject matter for being commercial in nature, and opted instead for DoD’s definition in its Other Transactions (OT) Guide for Prototype Projects.  In the absence of a clear statutory definition, GAO said it will defer to an agency unless the agency’s definition is “improper, ambiguous, or should be disregarded in favor of another definition.”  GAO declined to rein in DoD’s broad definition of “prototype,” which extends well beyond what many may consider the plain, common meaning of the word, to include “pilot” or “test” programs and even “demonstrations” of capabilities.  Specific to Oracle, GAO found that the term prototype also included the “demonstration of a repeatable framework of tools, processes and methodologies for securing, migrating (re-hosting) and refactoring, existing applications into a government-approved commercial cloud environment” that had been called for in the subject OTA.

Key Takeaways

GAO’s decision in Oracle offers some important and interesting takeaways for government contracting officers and contractors alike:

  1. Every DoD prototype OTA must include a provision expressly stating that DoD may award a follow-on OTA or contract for production work. This provision does not need to guarantee a follow-on award; it simply must provide that DoD may issue one.  But, bear in mind that an agency cannot provide for follow-on work in a prototype OTA awarded on a sole-source basis or otherwise without competitive procedures.
  2. Beware of substantial modifications to the scope of a prototype project. If a modification is in the best interests of both the contractor and the Government, the parties should agree, but the parties must be careful not to inadvertently delay transition of the initial prototype work into production.
  3. The universe of possible protesters that may challenge a follow-on production award is not limited to those who competed for the initial prototype OTA. To file a protest at GAO, one must be an “interested party,” which is typically “an actual or prospective bidder.”  Here, Oracle did not submit a response during the CSO process, but GAO nonetheless concluded Oracle was an interested party because it would be a prospective bidder for the follow-on production work were the work to be competed.
  4. The timing of an OTA protest must be carefully considered. A protest challenging an agency’s award of a follow-on production OTA or contract must be filed within 10 days of when the protester knew or should have known that the follow-on award had been made.  This is normally within 10 days of the publication of the notice of award, absent unusual circumstances.  However, GAO indicated in a footnote that, at least in Oracle, the protester could not challenge the award of follow-on production work by arguing that DoD had used inadequate competitive procedures to award the initial prototype OTA, indicating that such a protest would be an untimely challenge to the original solicitation.  For an argument to be untimely implies that it was timely at some earlier date, but GAO typically will not review competitions to award OTAs (except to determine whether a procurement contract should have been used instead).  Nevertheless, GAO in Oracle appears to be telling protesters that they must bring their challenges to prototype OTA competitions at the time of initial OTA award, perhaps indicating GAO’s willingness to consider such protests at least where the awarded OTA provides for follow-on production work.

Implications Beyond DoD:  Can Other Agencies Use OTAs to Acquire Goods or Services?

Over time, GAO’s decision in Oracle, coupled with its prior decision in MorphoTrust, may have greatest effect not within DoD but on the scope of other agencies’ OTA authority.  If GAO extends to other agencies the same standard of review it used in Oracle to review the scope of DoD’s authorizing statute, then other agencies may arguably have near-boundless authority to award OTAs for acquisition of virtually any product or service.

In MorphoTrust, GAO examined the Transportation Security Administration’s (TSA) decision to award an OTA for airline Pre-Check enrollment services.  TSA’s OTA authority, like that of many other OTA-authorized agencies, is more open-ended than DoD’s authority.  By statute, “The Administrator [of TSA] is authorized to enter into and perform such contracts, leases, cooperative agreements, or other transactions as may be necessary to carry out the functions of the Administrator and the Administration.”  49 U.S.C. § 106(l)(6); see 49 U.S.C. § 114(m)(1).

Prior to MorphoTrust, GAO applied the Federal Grant and Cooperative Agreement Act of 1977 (FGCAA), in addition to an agency’s authorizing statute, when reviewing an agency’s authority to enter into an OTA.  In relevant part, the FGCAA reads:

An executive agency shall use a procurement contract as the legal instrument reflecting a relationship between the United States Government and a State, a local Government, or other recipient when . . . the principle purpose of the instrument is to acquire (by purchase, lease, or barter) property or services for the direct benefit or use of the United States Government

31 U.S.C. § 6303(1).

Prior to MorphoTrust, GAO had sustained protests where agencies had attempted to acquire goods or services through an OTA rather than a procurement contract.  See, e.g., Rocketplane Kistler, B-310741, Jan. 28, 2008, 2008 CPD ¶ 22; Exploration Partners, LLC, B-298804, Dec. 19, 2006, 2006 CPD ¶ 201.  This was an important check on the agency’s authority, as most agencies do not have the same pre-award restrictions as those DoD must clear before awarding a prototype OTA.  See, e.g., 51 U.S.C. § 20113(e) (granting NASA authority to “enter into and perform such contracts, leases, cooperative agreements, or other transactions as may be necessary in the conduct of its work and on such terms as it may deem appropriate” (emphasis added)).

In MorphoTrust, however, GAO determined that the FGCAA “makes no mention of ‘other transaction’ authority” and therefore “provides no guidance when determining when an agency may properly use its other transaction authority.”  Without the FGCAA as a guidepost, GAO must look only to whether the agency’s decision to award an OTA was “knowing and authorized.”

This approach makes sense—even is necessary—for DoD’s prototype OTAs, given that by their nature (and as authorized by Congress) they essentially are acquisition instruments.  It does not translate, however, to other OTAs, which are by most accounts non-procurement instruments.  It is unclear whether GAO abandoned the use of the FGCAA in examining OTA authority when adopting MorphoTrust’s standard of review (considering “whether the agency’s use of its discretionary authority was proper, i.e., knowing and authorized”) in the Oracle case.  This seems unlikely, as just this April GAO updated its bid protest regulations to clarify that it will “review protests alleging that an agency is improperly using a non-procurement instrument to procure goods and services,” indicating the FGCAA’s legacy lives on.

Sometime soon, GAO likely will be asked to decide whether NASA (or some similarly authorized agency) has been unshackled from the FGCAA and may procure all kinds of goods and services through OTAs, free from the Federal Acquisition Regulation, the Competition in Contracting Act, and many other procurement laws.  The answer to this question surely must be “no”—but getting there may require either legislation or GAO making some distinctions that have not been readily apparent in its recent OTA-related decisions.

*Victoria Dalcourt is a summer associate in our Washington, D.C. office and is not admitted to the bar.