Some Shrinkage Questions

At the recently concluded Olympic Games, divers could earn a higher score if they performed tougher techniques. I'm going for my own high degree of difficulty, here: I plan to link technical accounting terms, the role of morality in American civic life, and some of the current election-year debates.

Let me know if I make a splash.

Start with one of those accounting terms: shrinkage. This is a line item in most retail corporate budgets. It's the jar of pickles that hits the floor, the pair of pants that walks out the front door, the hunk of fish that's left unsold because not enough people ordered the nightly special.

Depending on your theology, shrinkage is either a bean-counter's way of acknowledging the universal nature of Murphy's Law or a reflection of the essential Fallen nature of the world. In other words, shrinkage may be reduced but not avoided.

A good manager may tell the staff that the only acceptable goal for shrinkage is zero. But any CFO who fails to budget for a realistic level of shrinkage -- something well above zero -- deserves to be canned.

What is a realistic estimate of shrinkage for a large system? It varies widely, but I have an example. In last week's edition of the New Yorker, Dr. Atul Gawande continued his amazingly good series of articles about controlling the expenses of health care. In this one, he compared management in most hospitals to the well-controlled, high-quality operation of the Cheesecake Factory restaurants. Gawande mentioned that the chain aims to throw out no more than 2.5 percent of food due to over-buying.

So what about morality and civic life? If we are honest about the inevitability of shrinkage, we need to acknowledge the unavoidable imperfections in our civic systems. But where we allow for those imperfections to be expressed is not a bean-counter decision. It's like squeezing a balloon -- pushing down shrinkage here pushes other problems there.

To pull out another accounting term, shrinkage is fungible. That is, we can generally set a balance between two kinds of failures and decide which of the pair we are more willing to risk.

In theory, a grocery store could create a shelf where no jar of pickles ever falls. But that might make it so hard to a customer to get a jar that it kills sales. Cheesecake Factory could probably reduce that percentage of wasted food -- but only by disappointing more diners by being out of what they ordered.

America's Founders were totally aware of the limits of human systems. The most efficient form of government, after all, is a benevolent dictatorship. American democracy, with its multiple epicycles of checks and balances, is anything but efficient. Deciding how to balance the inevitable errors is a daily exercise in moral and political philosophy.

For instance, my high school civics teacher told me that the U.S. criminal justice system was set up with the value that it is better to let 100 guilty people go free than to convict one innocent person

That's a moral choice -- based on ideas of what is "best" that can't be captured on a spreadsheet. And it acknowledges the inherent error that is unavoidable in any large human system.

So on to examples from current politics.

Anybody who claims that eliminating "waste and fraud" will pay substantially for something in a government budget should be obligated to say just how much waste and fraud -- how much shrinkage -- is realistically acceptable. Anybody who answers "zero" gets canned along with my mythical CFO.

Take a mostly backburner issue: the death penalty. I have always wanted to ask proponents of the death penalty how many executions of the innocent they find to be acceptable. If they say "none," I want to ask what human system of any size has met that standard of perfection. Whether the realistic goal is one in 1,000 or 10,000 or 100,000, there will inevitably be human "shrinkage." Is it worth it?

On the other side, if the death penalty is a deterrent to murder, how many additional murders are acceptable if the death penalty is eliminated? Is that worth it?

Squeeze the balloon here and it pops out there.

The argument about voter ID is happening squarely on this morality-based playing field. On the one hand, how many qualified voters are you willing to disenfranchise? On the other hand, how many fraudulent votes are you willing to allow in? This is at least partly a bean-counter's calculation in that it surely looks like the voter ID laws will block many, many more qualified voters then they will stop what the evidence indicates is exceedingly rare ID-based ballot fraud.

But deciding the relative values of one disenfranchised voter compared with one fake ballot is a moral calculation, not simply a matter of stacking weights on each side of a scale. Far as I can tell, the GOP has thus far set the damage points for one fraudulent vote much higher than for a single legitimate voter denied a ballot.

Consider the balance-setting for social safety net and taxation issues. Do you want your shrinkage smaller where it might affect the basic needs of the poor? Or smaller where it might restrict businesses that can grow an economy that could eventually provide more jobs?

For gun control, is it a higher priority to keep weapons out of the hands of the wrong people or to make it easier for responsible people to buy a gun?

These are not questions with simple either/or answers, I realize. But our politicians speak too often as if they are -- with all of the downside risks placed on the side they oppose. I would love to see our candidates and politicians pushed to be more realistic in admitting the limits of the ideas they are proposing. Make them admit the inevitable likely costs, even as they tout the putative benefits.

I'm not holding my breath waiting for such honestly. After all, I'm asking the question "whose pain are you more willing to risk, and why?" That kind of frankness is as likely on the campaign trail as a cannonball dive at the Olympics.