A sunk costis an expense that already has been incurred and cannot be
recovered. For instance, in the earlier factory example, plant space origi-
nally may have been built at a high price. But this historic cost is sunk and
is irrelevant to the firm’s current decision. As we observed earlier, in the
case of excess, unused factory capacity, the relevant opportunity cost is
near zero.
More generally, sunk costs cast their shadows in sequential investment deci-
sions. Consider a firm that has spent $20 million in research and development
on a new product. The R&D effort to date has been a success, but an additional
$10 million is needed to complete a prototype product that (because of delays)
may not be first to market. Should the firm make the additional investment in
the product? The correct answer depends on whether the product’s expected
future revenue exceeds the total additionalcosts of developing and producing
the product. (Of course, the firm’s task is to forecast accurately these future rev-
enues and costs.) The $20 million sum spent to date is sunk and, therefore,
irrelevant for the firm’s decision. If the product’s future prospects are unfa-
vorable, the firm should cease R&D.
Perhaps the last word on sunk cost is provided by the story of the seventeenth-
century warship Vassa.When newly launched in Stockholm before a huge
crowd that included Swedish royalty, the ship floated momentarily, overturned,
and ignominiously (and literally) became a sunk cost.232 Chapter 6 Cost AnalysisBusiness Behavior:
Sunk CostsSunk costs are easy to recognize in principle but frequently distort decisions
in practice. The construction of nuclear power plants in the 1970s and 1980s
illustrates the problem. New plant construction was plagued by cost over-
runs and safety problems. (Indeed, after the Three Mile Island accident in
1979, safety concerns and strict safety regulations contributed to the overrun
problem.) At the same time, revenue projections declined due to the low
prices of alternative energy sources, oil and natural gas. While no newplants
were initiated in the 1980s (because of worsening profit prospects), many
utilities continued to spend on plants already in progress, despite equally
dim profit predictions. In light of uncertain profits and looming losses, mak-
ing the right decision—to continue construction or abandon the effort—
wasn’t easy. (As the unrepentant actress Mae West once said, “In a choice
between two evils, my general rule is to pick the one I haven’t tried yet.”) In
some cases, utilities abandoned plants that were 85 percent complete after
having spent more than $1 billion. Yet looking forward, this might be a per-
fectly rational decision. By contrast, construction of the Shoreham nuclear
plant on Long Island continued to completion despite severe cost escala-
tion and safety concerns. With an accumulated cost bill of $6 billion by 1989,
it never received regulatory approval to operate, and the enormous sums
spent came to nothing.c06CostAnalysis.qxd 9/29/11 1:46 PM Page 232