O.k. I wrote this years ago. Maybe 2003, maybe 2005:
And while NKs accept the basic logic of the rational expectations augmented Philips curve (that is, that the NAIRU is semi-fixed in the long run) monetary and fiscal policy can nonetheless be deployed in the short- run so long as the cause of the deterioration in effective demand is not caused by adverse supply shocks such as an unemployment rate below the NAIRU, insufficient capacity or medium to long-run supply constraints. This essentially amounts to a hawkish policy stance against inflation and support for less than full employment. Or alternatively stated, outside of a liquidity trap, NKs are almost indistinguishable in terms of macroeconomic policy from their new classical cousins.
 As we shall see below Shapiro and Stiglitz (1984) make the argument that unemployment (above its frictional level) is functional to aggregate efficiency.