The free market is a versatile tool for allocating resources.

So versatile that many consider it a panacea that solves all problems — not only in the public marketplace but also in business governance.

Transfer pricing (the practice formerly known as “charge-backs”) is just one example. Having an “invisible hand” make the complicated decisions required for properly allocating internally delivered services is tempting — much easier than making them through careful analysis and consensus building.

Sadly, free markets have their limitations (“externalities” in EconSpeak) — among them delayed-feedback-induced chaos, the “Tragedy of the Commons” (see “Laissez-faire Internetism,” IS Survival Guide, InfoWorld, 3/25/2006 for descriptions of both), monopolies and cartels, “network effects,” and the Jeffersonian notion of usufruct.

My friend Carlton Vogt pointed me to another — an economic conundrum called the dollar auction. Here’s how it works, as described by Professor Oliver R. Goodenough (Professor Goodenough’s article is about public policy — you’ve been warned):

I’m going to auction off a dollar bill. Every subscriber to this column can bid. The only difference between this auction and eBay is that the runner-up also pays.

So if someone else wins with a $0.50 cent bid and you were the runner up with a $0.45 bid, you’d send me 45 cents, the winner would send me 50 cents, and I’d send the winner the dollar bill.

Not wanting to spend $0.45 for nothing, you keep bidding, to $0.95. You’re the high bidder. The runner up is at $0.90 and has a choice: Lose $0.90, or up the bid to a dollar and break even. At which point your $0.95 cent bid is the runner up. But, you reason, a bid of $1.05 would cut your loss from the full 95 cents to just a nickel.

There is no logical end to this sequence — once you start, you’re trapped in the logic of ever-escalating investment whose sole goal is to cut your losses.

The dollar auction closely resembles the logic behind investing more time, money and effort in challenged business projects. It goes like this: “We’ve sunk so much money into this that pulling the plug would be political suicide. We’re better off biting the bullet and seeing the project through.” In dollar auction terms, it’s preferring to lose another dime instead of the entire last bid.

Sometimes, but not always, this logic makes sense — the extreme case is a project that needs one more week of funding. A business would be foolish to refuse it, simply because the project team missed the deadline.

The challenge governance committees face is recognizing the difference between dollar-auction situations and projects that deserve additional investment. Here are some indicators that you’re looking at a dollar auction:

• Goals: They no longer make business sense. Times change, situations change, projects stretch out. The idea has become obsolete, without ever having had a chance to be state of the art.

• Scope: When a project team has re-chartered itself with broader and more ambitious goals or deliverables, very often it’s because the team is failing to meet its deadlines. The broadened scope gives it breathing room.

• Plan: Look for holes — single tasks that will take a month or more to complete, tasks that have been 80% complete for more than two weeks, or worst of all, daily improvisation with no plan at all.

• Optimism: The project manager uses the word “hope” to describe any part of the plan. “Hope” is the work-breakdown-structure equivalent of the famous Gahan Wilson cartoon physicist whose mathematical proof included the step, “… and then a miracle happens.”

• No demo: If, by the time the question comes up, the project team can’t show working technology in action, it’s likely the technology will never work.

• Testing: If the project seems to be stuck in the testing phase … if the bug list isn’t shrinking on a weekly basis … the code, to use the technical term, “sucks.”

• Team: Project teams have to be fully committed to the project’s goals. This means project teams are required to lose their objectivity, and believe, deep inside, that the project will succeed. So if the project team recommends shutting things down, the situation has to be seriously awful.

No one indicator is definitive. The rating system:

1 or 2: Fix the issues, and go forward.

3 or 4: Worry, and conduct an external review before proceeding any further.

5 or more: The request for additional funding is a dollar auction, and the project is a zombie — the walking dead. Put it out of its misery.

Then burn the specs, just to be sure.