The Fed's Continuing Conundrum
New York (Nov 1) Coming soon to a central bank near you, in time for the Christmas shopping season, an increase in interest rates, courtesy of Janet Yellen and her colleagues on the Federal Reserve Board’s monetary policy committee. Or perhaps not. Folks living in euroland can expect the gift that keeps on giving, unless it doesn’t, more monetary stimulus. China works on a different calendar, so its people have already unwrapped their gifts, a cut in interest rates and assorted credit-easing measures. In Britain, the old naval slogan, “steady as she goes”, is likely to prove the best description of what the governor of the Bank of England has in mind. Which should all add up in the coming weeks to a strengthening dollar and less bounce in the step of exporters, inflation at only bit more than 1 percent, well below the Fed’s 2 percent target, and slowing growth. Inconveniently, just when Yellen & Co. have the last chance this year to consummate their flirtation with an interest rate increase, which they have hinted they still find attractive.
The Fed will have to balance economic and political considerations. It is under attack from politicians of both the right and the left. Republicans want to “audit” the Fed, and require it to set clear guidelines to follow when deciding on whether to raise rates, which most conservatives want it to do lest the currency become completely debased, if not immediately, then soon. Their real preference is to get rid of the Fed, return to the gold standard, and let the market set interest rates, but that is an agenda that can speak its name only in whispers and footnotes. Democrats want the Fed to keep the presses running lest the weak recovery be converted into another recession, and secular stagnation, the liberal hobgoblin since the days of the Keynesian revolution, becomes a permanent feature of our lives. The talk of an audit of Fed policymaking is prompting the Fed to lobby furiously to prevent what it sees as the politicization of monetary policy, an activity that Republican wannabee Rand Paul would have made illegal.
A dollop of sympathy for the Fed might be in order. As Philip Tetlock and Dan Gardner point out in their new, interesting book, “Superforecasting: The Art and Science of Prediction”, the average forecaster they studied over twenty years “did about as well as random guessing”, with the longer the look ahead, the less likely the forecaster to beat the accuracy of a dart-throwing chimpanzee.
Decades ago, as a night-school graduate student, I managed to get hired by an economic forecasting consultancy. The senior economists reserved to themselves the long-run forecasts, confident that they or their forecasts, or both, would be long forgotten when the actual data were compiled. I drew the short straw, and with modelling not yet in its infancy and the economy primarily a goods producer, manufacturing accounting for about half of all jobs compared with less than 20 percent today, decided to make four telephone calls each week: to General Motors to find out how many cars had rolled off the lines; to US Steel, to check on tons of output; to a leading carton manufacturer, to find the firm’s shipment of boxes to be filled with goods by manufacturers; and to the Atchison, Topeka and Santa Fe railroad for data on freight car loadings. The resulting forecasts were good enough to enable me to hang on to this much-needed job until I moved on to the next phase of my education.
Today, forecasters face a more difficult chore. Not only are they looking ahead through clouded windshields, but their rearview mirrors reveal an imperfect view of where the economy has been, as frequent non-trivial revisions to original reports demonstrate. And a service economy presents real measurement challenges, especially of productivity, the major source of rising living standards. We can measure with some confidence the number of man-hours it takes to turn out automobile grills, but are far less confident of our results when measuring the hours taken to invent and produce the software that accounts for about 25 percent of a new car’s cost.
Source: WeeklyStandard
