You can raise rates any time you want, but you can never leave the zero bound.

That, at least, has been the case for almost every country that has tried to increase interest rates from zero the last few decades. It hasn't been long until they've found themselves back where they began, at zero, with their economies still stuck in stall speed. Or, in Sweden's case, with even negative interest rates. (That means borrowers get paid to do so). The lesson is that raising rates too soon might mean that they are actually lower in the long run, since you will have to reverse course when you otherwise might have been ready to raise them.

Liftoff, in other words, isn't the kind of thing you want to get wrong. If you do, you might get stuck at zero for, well, who knows how long. It's been 15 years and counting for Japan since its first ill-advised attempt at hiking rates. That's a cautionary tale the Federal Reserve is well aware of as it contemplates raising rates as soon as this week, and maybe the most convincing argument for why it should hold off now. Think about it like this. If the Fed waits to long, inflation might get a little higher than it wants, but that's a problem it knows how to solve—with higher rates, which it also happens to want right now. If it goes too early, though, it might slow down the economy enough that positive interest rates become little more than an artifact from a bygone era.

But wait, why is raising rates from zero so difficult? It seems like you should just be able to ... raise them from zero. Well, there are two problems with that. The first is that an economy that needs zero interest rates is probably an economy that needs even more than zero interest rates. It probably needs negative interest rates, like the U.S. did, but can't get them since central banks can't cut rates that far without lenders hoarding money rather than paying people to borrow it. Now, it's true that central banks can make up for at least some of that by buying bonds with newly-printed money, but they don't like to do that. The result is that economies with zero interest rates don't get as much monetary stimulus as they need, so they don't grow as much as they could. And since rates follow growth, that means there isn't as much pressure for rates to rise once they do fall to zero.

The second reason lifting off is hard to do is that central bankers want to do it too much. They just don't like zero interest rates. Central bankers, after all, are supposed to be the ones taking away the punch bowl just as the party gets going, not plying recovering alcoholics with bottomless booze. Or at least that's what they tell themselves. That's why, when rates get to zero, central bankers have gone beyond looking for any excuse to raise them to inventing totally new ones to do so. The Bank of Japan, for example, hiked rates in 2000 even though core inflation, which strips out volatile food and energy prices to give us a better idea of the underlying trend, was negative, because, well, it's not clear why, but it had something to do with zero rates making things too easy for companies. Then it repeated this mistake six years later, when it once again tried to lift off despite negative core inflation. The European Central Bank didn't do much better when it let an oil price blip fool it into raising rates twice in 2011 even as other inflation was quiescent. And Sweden's central bank, the Riksbank, came up with an entirely new kind of error when it justified rate hikes first by pointing at inflation that wasn't there, and then by saying it was really about stopping a bubble.

Put these together, and it's easy to see why countries have had so much trouble escaping what economists call the zero bound. Remember, zero rates themselves mean that the economy probably isn't getting the stimulus it needs today. But by all-but-promising to end zero rates before the economy is ready for them to end, central banks are saying that it won't get the stimulus it needs tomorrow, either. That only slows down an already slowed-down economy even more, which, in turn, keeps rates low—as in zero. Now, there's a chance that the U.S. economy, with its steady 200,000 jobs-a-month growth, is strong enough to stay out of this trap. But there's also a chance that the U.S. economy, with its well below target 1.2 percent core inflation, isn't. Even worse, neither wages nor inflation have increased at all the past year. Raising rates to head off inflationary pressure you think will be there but isn't now is, as other countries have shown, a risky thing to do when rates are zero.

Welcome to the Hotel Zero Interest Rates might not be as catchy, but it's easier to get caught in.