The Federal Reserve is not an intelligence: it does not think. Individual members of the Federal Open Market Committee, however, do think. And the center of gravity of their individual thoughts now runs something like this:
1980-1985 was a disaster: 3 x 6 x 0.5 = 9%-point-years of lost jobs for Americans, jobs that would not have been lost had the Federal Reserve done its proper job and kept inflation under control in the 1970s:
Inflation exploded in the 1970s because of demand-pull: the Federal Reserve let employment get above full employment, and outgo workers and firms rapidly concluded that this was a permanent situation in which they had to act first to boost their incomes:
We cannot let this happen again: hence three times since 2010—in 2013, 2016, and 2018—we have used our policy tools of rates, quantities, and jawboning to push interest rates higher because we thought we were rapidly closing in on the full employment point at which inflation would start to move up substantially:
All three times we were wrong: we were not rapidly closing in on the full employment point at which inflation would start to move up substantially.
Now the inverted yield curve tells us that the market thinks that we have overdone interest rate increases, and will be substantially cutting rates by 100 basis points over the next two years:
The market could be right and it could be wrong. An inverted yield curve has been a reliable warning signal in the past, but the world is different now, and perhaps that matters. However, failing to begin to validate market expectations now would shake confidence and raise uncertainty. We have no warrant for believing that our models are more knowledgeable right now than the market. And there is pronounced asymmetry here: we can always deal with too-high inflation via tighter money, but it is not obvious what we could do to fix the situation if a negative demand shock were to push the prime-age employment rate down two or four percentage-points.
Hence we will validate market expectations and drop our policy rates by 25 basis points at our next meeting. What will follow that... will be data-dependent...