Krugman scolds the Fed. Unfairly.

So the Fed did not predict the panic in global stock markets.

So what?  Who could have?  It wasn't a mistake not to have.

Krugman doesn't scold them for this, but he does scold them for not postponing a hike on account of the asymmetric risks posed by stock market volatility.  The asymmetry being on account of the zero bound, which leaves them unable to stimulate if markets crash, but able to hike quickly if they boom.

But I don't think that the hike in and of itself shows that they failed to take account of this risk.

Think of a standard New Keynesian model projecting the awful shocks of the financial crisis, instructing the Fed to lower rates down to the zero bound, purchases assets [ok, I'm stretching the thought experiment here, but bear with me].   Eventually, once the effects of the shocks wear off, the model will be signalling that rates should be normalised.

The trajectory for rates will look different because of the asymmetry imposed by the zero lower bound [and the inability of asset purchases to make up for it].  From memory, it would involve a sharper cut on the way to the zero floor, and then a more protracted period of low rates [Woodford's 'lower for longer'].

But, at some points interest rates would rise.  So the Fed might be judged simply to have followed through on its asymmetry-inclusive path.

Or, we could think of them responding to positive news relative to the previous plan, which, again, zero bound included, would warrant a hike at some point, and this time earlier than before.

There are a few Fed speeches [I recollect Evans and Williams at least worrying about asymmetry] that cover the issue too, so this is more circumstantial evidence on which to acquit the Fed.

[Added later..]

Taking the argument to the extreme;  for a hike always to be wrong in the face of the zero bound implies that rates have to stay where they are forever.  Which is clearly not right.  So there is some prior forecast, or some incremental news that would have justified a hike, in the face of potential market volatility and the zero bound.  It's just a question of whether that hurdle was cleared in this case or not.