When the Federal Reserve first introduced its either/or stance on quantitative easing, I wasn’t sure if it was a PR ploy or a serious plan.

Specifically, the Fed said it was prepared to increase or reduce the current $85 billion-a-month pace of long-term asset purchases based on certain economic criteria. If the outlook for the labor market deteriorates, for example, the Fed might up the ante. Should hiring increase and inflation remain subdued, the Fed could taper its buying.

At his March 20 press conference, Fed chief Ben Bernanke said “it makes more sense to have our policy variable,” with purchases that respond to changes in the outlook.

To him, perhaps. If I understand Bernanke, he is saying that every six weeks policy makers will examine an array of leading, coincident and lagging indicators, most of which are revised and subject to seasonal distortions, to take the economy’s pulse and reassess the forecast. From there, they will determine the appropriate amount of monthly bond purchases.

This idea is as infeasible in theory as it is in practice.

Unlike the physician who uses real-time feedback to adjust the dose of a patient’s medication, central banks operate in a world of long and variable lags. Their predictive models have a poor record. The continuation of QE has always been predicated on an improvement in “the outlook for the labor market,” rather than an improvement in the labor market per se. Call it a better jobs market once removed. The inherent flaws in the theory should be apparent.

Methodological Mystery

As a practical matter, the plan is no more viable.

“The Fed doesn’t have a methodological way of calculating the relationship between asset purchases, interest rates and the economy,” says Jim Glassman, senior U.S. economist at JPMorgan Chase & Co.

He’s right. But you could say the same thing about the Fed’s traditional policy tool, the federal funds rate, and the preference for adjusting it in 25-basis-point steps, according to Neal Soss, chief economist at Credit Suisse Group AG in New York. Both have “the same element of science, judgment, and trial and error,” Soss says.

He’s right, too. But interest rates are a lot more visible. The public knows when the central bank raises its benchmark rate, even if its relationship to the rates on home mortgages, car loans and credit-card debt remains something of a mystery: It costs more to borrow.

On the other hand, the public doesn’t know, or care, about the size of the Fed’s balance sheet, reported every Thursday at 4:30 p.m. Even sophisticated traders and investors have trouble understanding the implications. They heard Bernanke say “exit” in congressional testimony two weeks ago when the Fed is ”still in the ’entry,’ or easing phase, of the policy cycle,” Soss says. A reduced pace of asset purchases still qualifies as stimulus.

An example will serve to demonstrate why using real-time data to adjust QE is a fool’s errand. As initially reported by the Bureau of Labor Statistics, non-farm employment averaged 157,000 a month in the fourth quarter. Subsequent revisions raised the average to 209,000.

Last week, the Bureau of Economic Analysis uncovered an additional $108 billion (annualized) of personal income in the fourth quarter, which had to come from a larger workforce, higher compensation or some combination of the two. The revision was based on the Quarterly Census of Employment and Wages, an almost-universal tally of jobs and wages derived from tax reports submitted by employers.

The jump in compensation was partly a function of income-shifting to avoid the year-end tax increase. When the report is released to the public on June 27, the source of the understatement will become clear. With its intense focus on the labor market, would the Fed have made the same modification to QE six months ago, based on available information, as it would today? Somehow I doubt it.

Feedback Loop

Variable QE would represent an extreme case of micromanagement that is likely to run amok. Economists are already starting to forecast the size and timing of reductions to QE. That means the markets will build an expectation into prices, which means the Fed will have to consider those expectations in deciding what to do, which means a closed feedback loop that is of no value to anyone.

The bond market sell-off in response to Bernanke’s tapering talk is a taste of what’s to come if the Fed decides to implement variable QE. Market volatility will swamp any perceived benefits. That’s why there’s a greater risk of the Fed overstaying its welcome with accommodative policy than withdrawing it too soon.

Which brings us to the possibility that this may all be a public-relations gimmick designed to prevent financial markets from getting four steps ahead of the Fed: from $85 billion a month in bond purchases to the termination of QE to outright sales from its portfolio to an increase in the funds rate. Such a ploy would run counter to Bernanke’s DNA. After elevating communication to a policy tool, the last thing he would want to do is mislead the markets.

So which will it be? An option that is unworkable or one that would undermine the Fed’s credibility? It sure sounds like a choice between two losing propositions.

(Caroline Baum, author of “Just What I Said,” is a Bloomberg View columnist. The opinions expressed are her own.)

To contact the writer of this article: Caroline Baum in New York at cabaum@bloomberg.net

To contact the editor responsible for this article: James Greiff at jgreiff@bloomberg.net