On Friday afternoon, the Government Accountability Office released its full legal decision in the ENCORE III case we reported on a month ago. That’s the case in which GAO agreed with two large vendors that that DoD’s $17.5 billion solicitation for IT services was flawed and needed to be rewritten.
The decision sheds a lot of additional light on the reasons for GAO’s decision, and interestingly, as the contract arbiter sees things, the main problem with ENCORE III wasn’t strictly that the Defense Information Systems Agency wanted to award spots on the ten-year contract on a lowest-price technically acceptable (LPTA) basis, a complaint industry groups had been making for months leading up to the release of the final request for proposals.
Indeed, GAO explicitly rejected the claim that the agency shouldn’t have used LPTA, saying the decision was justified because ENCORE is “a mature program with a substantial commercial application.” The problem was that DISA contradicted itself in another part of the multiple award ID/IQ contract when it said that even though it wanted labor rates to be as low as possible, vendors would be automatically shut out of the competition if their rates were too low – a notion that seemed to be at variance with the common understanding of the word “lowest.”
So, exactly how low was too low? It was impossible for vendors to know at the time they placed their bids. The methodology DISA had planned to use was to take all the proposals they received and average them together – leaving out the 10 percent of the most expensive vendors as well as the prices in the bottom 10 percent.
From there, the agency would use that “trimmed average” to automatically disqualify any company whose prices were either 50 percent above or below the mean.
Procurement officials said they had a perfectly sensible reason for doing this, based on their experience with ENCORE III’s predecessor, ENCORE II: choosing 20 vendors whose prices weren’t too high or too low would help ensure that they stood a chance of winning the task orders to follow, and that DoD wouldn’t get stuck with companies who’d lowballed their bids just to get onto the contract but couldn’t actually perform the work.
But GAO noted that the Three Little Bears approach (my term, not theirs) to source selection wasn’t compatible with the idea of “lowest price,” and more importantly, that the thresholds DISA planned to use to cut off vendors based on price were completely arbitrary. That’s partially because they were based on the track record of ENCORE II, which wasn’t an LPTA competition.
“While we certainly agree with DISA that agencies enjoy broad discretion when establishing their bases for evaluation, their selected evaluation method must have some rational basis,” GAO wrote in the protest decision. “Here, the record is devoid of any basis to conclude that the 50 percent price floor for the mechanical elimination of proposals reflects a point at which there is a particularly high performance risk to the government…the agency has been unable to articulate, or otherwise provide evidence to support a conclusion regarding how risk increases to an unacceptable level.”
GAO explicitly rejected the idea that agencies can’t use LPTA for ID/IQ contracts that demand many years of complicated technical services, a precedent that companies are likely to be unhappy with.
But besides making clear that agencies need to have a rational basis for evaluating various firms once they’ve gone down the LPTA road, the decision signals that in big IDIQs, the government will need to give companies a better idea of what it intends to buy under cost-reimbursable task orders.
The ENCORE III RFP includes both cost-plus and fixed-price line items, and GAO found that DoD was asking companies to compete for ENCORE without giving them a chance to explain why the proposals they offered on the cost-reimbursable parts of the contract outshone their competitors, since DISA didn’t even ask for proposed prices.
“The solicitation does not seek any information for, or provide for the evaluation of a segment of orders that may represent half of the [ENCORE III] requirement. It fails to provide an adequate basis to compare the relative cost to the government of the competing proposals,” GAO attorneys wrote.