Sunk Cost

In economics and business decision-making, sunk costs are retrospective (past) costs that have already been incurred and cannot be recovered. Sunk costs are sometimes contrasted with prospective costs, which are future costs that may be incurred or changed if an action is taken. Both retrospective and prospective costs may be either fixed (continuous for as long as the business is in operation and unaffected by output volume) or variable (dependent on volume) costs. Note, however, that many economists consider it a mistake to classify sunk costs as "fixed" or "variable." For example, if a firm sinks $1 million on an enterprise software installation, that cost is "sunk" because it was a one-time thing and cannot be recovered once expended. A "fixed" cost would be monthly payments made as part of a service contract or licensing deal with the company that set up the software. The upfront irretrievable payment for the installation should not be deemed a "fixed" cost, with its cost spread out over time. Sunk costs should be kept separate. The "variable costs" for this project might include data centre power usage, etc.
In traditional microeconomic theory, only prospective (future) costs are relevant to an investment decision. Traditional economics proposes that economic actors should not let sunk costs influence their decisions. Doing so would not be rationally assessing a decision exclusively on its own merits. Alternatively, a decision-maker might make rational decisions according to their own incentives, outside of efficiency or profitability. This is considered to be an incentive problem and is distinct from a sunk cost problem.
Evidence from behavioral economics suggests this theory fails to predict real-world behavior. Sunk costs do, in fact, influence actors' decisions because humans are prone to loss aversion and framing effects. In light of such cognitive quirks, it is unsurprising that people frequently fail to behave in ways that economists deem "rational."