That’s because there’s less certainty today over the direction of Federal Reserve policy than at any time since late 2012, when Ben Bernanke launched his controversial third round of quantitative easing. It’s widely expected that Yellen will emerge from the deliberations on Wednesday announcing that the Fed intends to keep short-term interest rates near zero. But what comes after that, at the Fed’s meeting in January and beyond, is anyone’s guess.

The Fed’s behavior—hinting at rate increases but never delivering—has even prompted economist John Taylor, a monetary policy expert, to assert that, “nobody knows what [the Fed] is doing.” In an age when the Federal Reserve is communicating more and more with the public, the public seems to have less and less of an idea what the Fed is going to do next. And there’s a simple explanation for this confusion: economists have a much looser grasp on what causes changes in inflation and employment than they’d like us to believe.

Economist Milton Friedman proved that, over the long run, “inflation is always and everywhere a monetary phenomenon.” He compiled data that showed that the rate of inflation correlates strongly with the amount of money in circulation. But in the short run, the economy doesn’t always cooperate with this theory. Case in point: the Fed has recently flooded the economy with bank reserves and yet this hasn’t led to banks creating, through loans, the kind of money that is used by consumers and businesses.

This has left the Federal Reserve with what former Fed Chair Ben Bernanke called last week, “really the only model economists have” for predicting inflation, known as the Phillips Curve. Named after economist William Phillips, it describes the inverse relationship between employment and inflation. Because labor costs make up a large portion of the cost of goods and services, it would make sense for prices to rise as competition for workers becomes more intense.

But the empirical evidence for the Phillips Curve is tenuous. This Wall Street Journal chart shows the relationship between inflation and the unemployment rate over the past 50 years:

Neither Friedman nor Phillips’ theories can explain why inflation is so low in today’s world. Absent hard empirical data for what the FOMC should do, its members appear to be falling back on a combination of gut instinct and political bias to guide their decisions.

Take, for instance, a recent interview with Fed Governor Daniel Tarullo, in which he argued that delaying an interest rate increase is about risk management. He described himself as, “being concerned that a premature rise might be harder to deal with than waiting a little bit longer.” Tarullo doesn’t give any empirical evidence for this bias against raising rates; he just believes that the risks of higher rates killing the recovery are greater than runaway inflation. On the other side of the spectrum is the FOMC’s resident hawk, Governor Jeffrey Lacker, who argues that the risk of inflation spiraling out of control, as it did in the 1970s, is what we should be afraid of today.

Lost in the din of all these speeches, interviews, and board statements is an empirical foundation for the Fed’s decisions. This is not an indictment of the Fed per se. The central bank shouldn’t claim to have knowledge that it lacks. But it also shouldn’t be surprised that it’s failing to convince markets of its intentions when the theoretical basis of those plans are so shaky.