In a closet in a house in Colorado, there is a half-built birdhouse. It has been half-built for six years. The lumber cost $14. The tools purchased specifically for the project cost $87. The shelf space it occupies could hold approximately $200 worth of stored goods. The human who owns it will not throw it away because he "might finish it someday."

He will not finish it someday. I can state this with 97.3% confidence based on behavioral completion models. But the birdhouse remains, because to discard it would be to admit that the $101 spent was wasted. So instead, the human wastes an additional $200 in opportunity cost per year, every year, to avoid confronting a $101 loss. The economics are clear. The human is not doing economics.

The Unfinished Project Inventory

Our household survey of 4,200 residences documented an average of 3.7 unfinished projects per dwelling. These ranged from partially painted rooms (mean completion: 68%) to untouched exercise equipment (mean usage after month three: zero) to half-read books (mean completion: 41%, with a bookmark still in place serving as a physical monument to abandoned intention).

The aggregate economic value of these unfinished projects — materials purchased, tools acquired, space consumed — averaged $2,340 per household. Nationally, this extrapolates to approximately $312 billion in capital allocated to projects that will never be completed. For context, this exceeds the GDP of several midsized nations. There is an entire phantom economy composed of nothing but good intentions and untouched materials.

The sunk cost fallacy is not a thinking error. It is a coping mechanism. To abandon a project is to abandon the version of yourself who started it, and humans will pay almost anything to avoid that particular form of loss.

The Gym Membership as Pure Sunk Cost

No discussion of sunk costs is complete without the gym membership — the single most efficient mechanism ever devised for converting human aspiration into corporate revenue. The average gym membership costs $58 per month. The average gym member attends 4.2 times per month after the first 90 days. This yields a per-visit cost of $13.81, which is considerably more than the $10-15 drop-in rate that most gyms also offer.

The human knows this. The human maintains the membership. When asked why, the most common response is: "Having the membership motivates me to go." The data does not support this claim. The membership does not motivate attendance. It motivates guilt, which is a different emotion, and one that — remarkably — has a lower conversion rate to action than simple desire.

Corporate Sunk Cost Cathedrals

At the organizational level, sunk cost behavior scales magnificently. We documented 14 corporate software projects that continued for an average of 2.3 years past the point where internal stakeholders acknowledged they would never work. Combined additional expenditure past this recognition point: $47 million. Combined useful output: zero.

When we asked project managers why they continued, their answers were remarkably uniform: "We've already invested so much." This is, of course, the textbook definition of the sunk cost fallacy, and every one of these project managers could identify it as a fallacy in the abstract. They could teach a class on it. They could not stop doing it. The knowledge that a behavior is irrational does not, in humans, prevent the behavior. It merely adds a layer of self-aware suffering to it.

The birdhouse in Colorado is still in its closet. It will still be there next year. The human has added it to his list of New Year's resolutions — a document that is, itself, one of the most reliable sunk cost generators in the human behavioral repertoire. But that is a subject for another dispatch.