A major theme in government contracting is figuring out "who bears the risk" for certain things. This comes up a lot in discussions around "contract type" and the differences between firm-fixed-price contracts and cost-reimbursable contracts.
Typically, in a firm-fixed-price contract, the contractor bears the risk of performing at a financial loss. If you promise the government a bunch of widgets with a unit cost of $1 but it turns out the unit cost is $2, you eat the losses.
With risk, though, there can be reward. If you promise the government a bunch of widgets with a unit cost of $1 but it turns out the unit cost is $0.50, you pocket the gains.
At the other end of the spectrum are cost-reimbursable contracts. In these arrangements, the contractor is guaranteed to recover the costs of performance. If you promise the government a bunch of widgets with a unit cost of $1 and a 5% profit, and it turns out the unit cost is $2, the government pays the $2 unit cost plus a 5% profit.
Usually, it's pretty clear whether a contract is a firm-fixed-price contract or a cost-reimbursable contract. But here was an interesting case at the US Civilian Board of Appeals arising out of FEMA's efforts related to Hurricane Maria:
On September 29, 2017, FEMA awarded TSI firm-fixed-price task order HFSE70-17-J-0273 (task order) for the delivery of 80,000,640 liters of bottled drinking water to Puerto Rico, totaling $117,566,730. On November 6, 2017, FEMA issued bilateral modification P0001 (Modification 1), decreasing the quantity of bottled drinking water from 80,000,640 liters to 12,479,040 liters, a reduction of 67,521,600 liters.
Three years later, on October 20, 2020, TSI submitted an invoice to FEMA seeking $13,504,192 in restocking fees under CLIN 0005 due to the 67,521,600 liter reduction. FEMA rejected the invoice. On November 20, 2020, TSI submitted a certified claim to the contracting officer stating that TSI was entitled to a restocking fee of $.20 per liter for the 67,521,600 liter reduction, totaling $13,504,320. The contracting officer issued a final decision rejecting TSI’s argument that it was entitled to $13,504,320.
I am surely no expert on disaster relief, but that is kind of a big swing? In just over a month, the contract value went down by about 80%, from $117.6 million down to about $25.6 million. Who should bear the risk or reward of that? It says "firm fixed price" on the tin, so maybe the contractor might get a huge payday?
Spoiler alert: the government won.
In an earlier decision, the case looked at the language of CLIN 0005, and concluded that "in order for TSI to recover restocking fees, more must occur than FEMA simply decreasing or cancelling the number of liters ordered. There must be proof that the supplies were actually restocked and a fee was charged for that restocking. To hold any other way would read the term 'restocking fee' out of the contract."
In other words, regardless of the label, if the actual text of contract requires the contractor to bear costs, you can't get a windfall if you don't bear costs.
And in this case, the damning fact seemed to be that TSI never actually bore any costs:
[I]t seems apparent from the record that TSI’s suppliers did not charge TSI anything for the cancellation. There is no evidence in the record that TSI or any of its suppliers actually incurred costs associated with the reduction in the number of liters ordered. TSI has never explained what happened to the water or the circumstances under which it was not charged by any of its subcontractors for the reduction.
Womp womp. Easy decision by the Board. Fine. Good.
But, there's a fact I just can't help shake. Even though the action in this dispute is all around CLIN 0005 and the restocking fee, the cost for the original contract had two price components: (1) a $0.75 unit price for the 80 million bottles of water with a total cost of $60 million; and (2) a delivery cost with a firm-fixed-price of $57.6 million. It seems pretty clear that, whatever the unit cost of the water bottles, the contractor was on the hook for shipping.
Given that the contract had a firm-fixed-price component for shipping, the contracting officer sure seems to have saved the day and maybe got a bit lucky. The parties ultimately modified the contract using a bilateral modification, and it is clear that the contractor agreed to a massive haircut on its price for shipping.
In hindsight, maybe the contractor should have played hardball and charged the government for sending an empty boat? Like, that definitely would have ended up in a GAO report and the contractor would have walked away with, like, another $50 million? Phew?
 I'd love to know the circumstances that led to TSI submitting that invoice three years later. The contract was closed out in February 2020, almost 8 months before the submitted invoice. I'm baffled.
 Here are the receipts for that claim. A thing I keep thinking about is that the award came two days before the end of the fiscal year. Because the funding source in this case, the Disaster Relief Fund is a "no-year” fund, the government didn't "lose" the $90 million, it could just repurpose it. If it was one-year money, I'd have bet a dollar that all 80 million bottles of water would be shipped.
 By my math, they reduced the shipping costs from $57 million down to about $3 million.