SINCE interest rates hit rock-bottom in 2009, the Federal Reserve has repeatedly made optimistic forecasts about when they would start rising, only to delay the big day again and again. If the Fed has been a bullish coach, the markets have been trusting fans, continually believing that an increase is imminent, only to have their expectations dashed. At last, however, the moment seems to have arrived. On December 16th, when the Fed’s rate-setting committee meets, it seems all but certain to raise rates.

For that, thank the strength of the labour market. Unemployment, at 5%, is as low as most analysts reckon it can sustainably fall. During the recession, America lost 8.7m jobs. It has since gained 13m. In 2010 there were six unemployed workers for every job opening; today there are 1.5.

Wages, long stagnant, finally appear to be growing again, too. In September, when the Fed toyed with raising rates, average hourly pay had just grown by 2%, on an annualised basis, over the prior three months. Now that has risen to 2.8% (see chart). By one measure, wages grew by fully 4% in the third quarter of the year. Accelerating pay suggests that slack in the labour market has almost gone.

The pick-up in wages could peter out, however. Since the recession hordes of workers have left the labour force altogether: labour-force participation among 25- to 54-year-olds in the third quarter of the year was lower than at any time since 1984. If some can be tempted back to work, wages will be held down.

Moreover, the Fed’s preferred inflation measure stands at just 0.2%, well below its 2% target.

Cheap oil bears much of the blame. But core inflation, which excludes volatile energy and food prices—and so is a better indicator of underlying price pressures—is only 1.3%.

Janet Yellen (pictured), the Fed’s chair, chalks up some of the shortfall to a strong dollar making imports cheap (the greenback is up by 19% since mid-2014). That effect should dissipate if the dollar’s ascent stops. There is also less scope for the oil price to plunge, having already fallen by almost two-thirds over the past year. This suggests inflation may pick up in 2016. That, in turn, argues for a rate rise soon, since monetary policy is thought to have only a delayed impact on the economy.

The Fed is in a jam, though, because it faces asymmetric risks. If it raises rates too soon, its scope to cut them, should the economy then sour, is limited by the fact rates cannot fall far below zero. If it waits until inflation is stronger, it has unlimited capacity to raise rates to tame it.

Getting this balance right will be tricky. Ms Yellen likes to emphasise that starting early keeps the journey smooth; abrupt rises later might rattle markets. Yet the Fed’s forecast for rates is steeper than what the market predicts. The Fed’s median rate-setter expects interest rates to rise to around 1.5% by the end of 2016; by contrast, traders expect rates of only 0.85% in a year’s time.

Two roads in a Wood
Who is right depends on how the economy reacts to the first rise. That is hard to predict, because the channels through which monetary policy works are mysterious. Take consumption, which has driven America’s recent growth. The impact of rates on consumer spending is muted by the fact that, unlike in much of Europe, most American mortgages come with fixed interest rates, shielding many Americans from swings in monetary policy. The first rate rise will nudge up the cost of borrowing, but only very slightly.

Nor is a rate rise likely to slow investment much. The evidence for the responsiveness of investment to rates is mixed; business confidence is probably more important. If—as some think—a rate rise is a signal from the Fed that America’s economy is healthy, investment could even rise.

That leaves exports. If rising rates cause the dollar to appreciate further, American goods will become still more expensive abroad. America’s embattled manufacturers will not welcome that (a recent survey suggests manufacturing output shrank in November for the first time in three years). But another surge in the dollar is unlikely, since a rate rise in December is now widely expected.

If the Fed follows through on its forecasts, though, and raises rates faster than markets expect in 2016, the dollar may well rise further, dampening inflation quickly. Stanley Fischer, the Fed’s vice-chair, recently estimated that a 10% rise in the dollar reduces core inflation by half a percentage point within six months. For all her horizon-gazing, Ms Yellen is unlikely to persist with rapid rate rises if they push inflation too far below target in the short term.

Further falls in commodity prices could also keep the brakes on. This would drag down inflation directly, but could also reduce it indirectly by pushing up the greenback. Much of the dollar’s recent appreciation, on a trade-weighted basis, derives from weakness in the Mexican peso and the Canadian dollar. Those currencies weaken when commodity prices fall, argues Paul Ashworth of Capital Economics.

Most uncertain of all is where interest rates will end up. That depends on the so-called “natural” rate of interest; the sweet-spot which balances demand and supply. This is tricky to pin down, but it is commonly thought to be falling, in part due to systemically slower growth since the crisis. One estimate—based on work by John Williams, a rate-setter himself—puts the inflation-adjusted natural rate of interest at -0.1%, down from 3.1% in 2000. Given an inflation target of 2%, that points to rates eventually settling at just under 2%. That is worryingly low; in 2007, before the crisis, the Fed had leeway to cut rates by over 5 percentage points.

The last time monetary policy changed in a comparable way was in 1947, when the Fed started raising rates from a lowly three-eights of a percent, where they had sat for five years. This time, the wait for another stint near zero may not be nearly so long.