This would be quite the paradox. Lower prices at the pump, after all, are supposed to help the economy by putting more money in people's pockets.
Why does the IMF think otherwise? Well, it has to do with what economists call the real interest rate, which is just the same old interest rate everyone talks about minus inflation. That, you see, is what determines how much businesses do or do not want to invest. Think about it like this: you're going to be much more willing to pay 5 percent to borrow if you believe prices are going to rise 5 percent themselves than if you believe they aren't going to rise at all. That's because your income should go up with inflation, but your debts won't. The most important word here, though, is "believe". It's inflation expectations more than inflation itself that matter when you're trying to decide whether to borrow money for a few years, or, in the case of a mortgage, decades. But in any event, the big picture is pretty straightforward: The lower the real interest rate is, the more people will tend to invest, and the more people invest, the more the economy will tend to grow.
There's one more wrinkle, though. That's the fact that interest rates can't go too far below zero. If they did, then people would just pull their money out of the bank rather than pay a fee to keep it there—that's what a negative interest rate would be—and stuff it in their mattress instead. So in this world where everyone was holding their money in cash, the interest rate on cash would become the interest rate for the whole economy. What's that? Well, zero. Cash, after all, doesn't pay any interest, but there's no way it could make you pay a penalty, either. And that's why economists say there's more or less a "zero lower bound" on interest rates.
Now, if you think about it, this means anything that reduces inflation when interest rates are zero would also reduce growth. That's because, in that case, lower inflation would mean higher real interest rates. Those, remember, are just regular rates minus inflation. So if the number you're subtracting (inflation) gets smaller, but the number you're subtracting it from (interest rates) stays stuck at zero, you'll get a bigger answer. And in the extreme, these higher inflation-adjusted interest rates—the borrowing costs that matter—could even slow the economy down more than lower prices at the pump speed it up.
The question, then, is whether lower oil prices would really turn into lower inflation expectations. If they did, then cheaper oil might be bad for the economy right now by raising real interest rates. And the flip side of that is that more expensive oil might be good by theoretically increasing inflation expectations and reducing real interest rates. It'd certainly be, as Obstfeld puts it, a "perverse" result.
Could this really be true? Well, it is the case that strange stuff can happen when interest rates are zero. The simple story is that the way the economy works doesn't work automatically. Something has to change to make it work—and that something is usually interest rates. In other words, the economy depends on interest rates adjusting to whatever's happening, with the operative word being adjusting. Consider this: the reason new savings get turned into new investments and new job-seekers get turned into new jobs is that the first half of each of these pairs reduces rates, and those lower rates make the second half happen. So if rates can't fall—because they're at zero—then the economy won't function like it normally does. It won't function at all.
But just because this could be true doesn't mean it is. The only way to tell, as we mentioned before, is if lower oil prices do turn into lower inflation expectations. If they don't, then there wouldn't be any paradox at all.
Now, this is where things get tricky. It seems obvious that lower oil prices should make people expect less inflation, but that's not how things are supposed to work. A one-time change like a fall in oil prices isn't supposed to change what people think about inflation in general. It's the difference between prices that go through a lot of up-and-downs and ones that don't. The first kind are hard to predict, so people don't try to when they, say, negotiate a raise. Instead, they look at prices that only change every now and then—i.e., not oil—to come up with their own inflation expectations. And the IMF economists, for their part, don't disagree. At the same time, though, they say that lower oil prices should mean lower prices for things that take a lot of oil to make, and that it's those declines that should push inflation expectations down. In other words, the second-order effects of falling oil prices are what might make inflation expectations fall too.
And it sure seems like that's what's going on. Oil prices and market-based measures of inflation expectations, which look at the difference between unprotected and inflation-protected bonds, actually have moved in tandem the past two years. You can see that clearly enough in the chart below from the IMF.
Or can you? The problem, as University of Oregon economist Tim Duy points out, is that this chart "really proves nothing." Sure, there's a correlation between oil prices and inflation expectations during this time, but there's also been one between the dollar and inflation expectations. Which one is the real story? Probably neither. Indeed, this relationship between oil prices and inflation expectations vanishes if you look at it over a longer period.
Not that this should be too surprising. Despite what the IMF might say, there's very little reason to believe that higher oil prices, for example, would really turn into higher prices for other things. At least not anymore. "Evidence of a pass-through effect," Fed researchers have found, "appears strong during the oil shocks in the 1970s, but seems to have disappeared since the mid-1980s." That, not so coincidentally, is around the time that the Fed stabilized inflation expectations. So it's the fact that the Fed made it clear it would do whatever it took to keep inflation from getting out of control that seems to have broken the connection between oil prices and inflation. Or, as the Fed economists put it, "oil price shocks appear to have only transitory effects on headline inflation and virtually no impact on measures of underlying trend inflation."
If anything, higher oil prices seem to cause more damage when interest rates are zero than when they're not. That, at least, is what University of California-San Diego economist Johannes Wieland found when he looked at Japan the last 30 years. Now, in a lot of ways, this is the perfect test case. Japan has had zero interest rates a lot longer than anyone else, and, since it imports all its oil, the only way more expensive crude could help it is through expectations. But that's not what happened. Oil shocks made unemployment, industrial production, and consumer spending all do worse when interest rates were zero than at other times. In other words, higher oil prices were always contractionary, but they were most contractionary at the zero lower bound. The only hint of a silver lining is that higher crude prices did seem to cause a very slight increase in inflation expectations when rates were zero. But even then, that's still consistent with the idea that there'd be little to no change.
That means that, even today, higher oil prices wouldn't really increase inflation expectations, wouldn't really cut real interest rates, and really wouldn't help the economy. All they would do is suck more money out of consumers' pockets. "It is absurd," Adam Posen, the president of the Peterson Institute for International Economics and a former member of the Bank of England, told me, "to think that higher oil prices would be expansionary for the overall economy—no matter how nice they might be for exploration companies in North Dakota."
Everything might not be as it seems when interest rates are zero, but that doesn't mean nothing is.