It's beginning to feel a bit awkward. Federal Reserve Chair Janet Yellen wants inflation to move back to 2 percent, yet every time oil takes a plunge, she tells us it doesn't matter. We first heard Fed policy makers describe the impact on inflation from oil's nosedive as "transitory" in December 2014, after West Texas crude had already dropped 42 percent from a year earlier. Crude has since pushed down another 44 percent and with every Federal Open Market Committee statement and every Yellen press conference, the Fed has stuck to its guns. Pay no attention to oil, it's "transitory." How long can Fed officials keep this up and remain credible? As long as oil keeps falling, actually. That's because when it comes to its relationship with inflation, oil is different. Prices for global commodities such as oil are mostly determined by current supply and demand, not the cost of production.  While there is some relationship, a producer can't easily adjust prices today based on costs. That differs from other goods and services — cars, for example —in which costs and prices are intimately linked. When labor costs rise for a car maker, they're usually passed on to consumers, which can in turn motivate broader wage pressure, feeding a cycle. That's what makes increases in the price of cars more meaningful for inflation than the price of oil. "Inflation is driven by a cycle of price increases leading to cost increases leading to price increases," said Barry Bosworth, a senior fellow at the Brookings Institution in Washington. That dynamic can occur in the vast majority of industries in the U.S., but not by agricultural and energy commodities. "When oil rises or falls, it doesn't have a persistence," Bosworth says. "It's a shock." In other words, it's still unlikely to instigate a broader and persistent cycle of downward price adjustments. A year after any move in the price of oil, its impact on 12-month inflation will mostly vanish. Michael Feroli, chief U.S. economist at JPMorgan Chase and Co. in New York, puts it another way, focusing on whether the price for a product or service is "sticky." "For things that have infrequent price adjustments, when those adjustments are higher, that probably reflects a forecast by the price setter that overall price pressure are going up," Feroli said. "So, sticky prices are the ones that are more informative" when it comes to predicting future inflation. And oil prices are not sticky. That's why economists, when they want to put their finger on inflation's true trend, exclude energy and food prices. That gives us "core" inflation and all it variations, including the Atlanta Fed's Sticky Price CPI, which filters the components of the consumer price index according to the frequency of their price adjustments. Twelve-month Sticky Price CPI was 2.4 percent in December, compared to 0.6 percent for the Fed's preferred gauge of inflation. So even if oil takes another swoon, economists, including Yellen, will still want to ignore it at least as far as inflation is concerned. There's just one big caveat, said Jeffrey Fuhrer, senior policy adviser at the Boston Fed. Oil can have a more lasting effect if it begins to influence inflation expectations — a crucial, if mysterious, element in what generates price pressures. Given oil's volatility, that really shouldn't happen. Yet, over the last decade, long-term inflation expectations — as measured by the compensation Treasury investors demand for inflation five to 10 years out — have shown a surprising positive correlation with oil prices. "This is a little bit of a puzzle to me," says Fuhrer. "To the extent this does happen, is the financial system aware this is a phenomenon likely to reverse, and will they move their expectations in reverse? My hope is they would."