"Based on the relatively strong economy, but the continued stubbornness of inflation, I've penciled in two hikes for 2018," Harker said at Saint Louis University in St. Louis, Missouri. "I use pencil because the data can change, and sometimes they don't accurately point to future events."
The Fed is widely expected to raise interest rates next month, the first of what many at the Fed believe should be three rate hikes this year. A recent strengthening in inflation data has helped convince many in financial markets that the Fed will need at least three rate hikes in 2018 to prevent the economy from overheating.
In his prepared remarks, Harker made no mention of the recent inflation data, nor of the tax cuts that some view as fueling faster price rises.
Sticking closely to a view he laid out earlier this year, Harker said he expects the U.S. economy to grow 2.5 percent this year before slowing to 2-percent growth next year and to below 2 percent in 2020.
Unemployment, he forecast, will fall from 4.1 percent now to 3.6 percent by the middle of next year before rising back up a few tenths of a percentage point, while job growth will remain strong.
And inflation, while still below the Fed's 2-percent goal, should meet or exceed that objective by the end of 2019, he forecast.
"The Feds mantra is data dependent, and for now, the data continue to tell me two (rate hikes) is the likely appropriate path," said Harker, who does not have a vote on the Fed's policy-setting committee this year but who takes part in the panel's regular meetings.
Harker said he is open to a rethink of the Fed's policy framework, but that he is in "no rush" to adopt, say, a higher inflation target or a nearly untested strategy, favored by a few of his colleagues, that would allow inflation to run hot for a period of time to make up for a period of excessively low inflation.
The central bank would need to think "long and hard" before making any changes that could put it out of step with other central banks globally, he said.