Return of the Old Keynesian Phillips Curve?

8/25/2013 12:48:00 PM
Paul Krugman posts an interesting chart of a Phillips curve relationship over the "Great Moderation" through today. I've reproduced his chart below, with the same data from Fred:

Notice I've highlighted the years that were due to the Great Recession, which has actually made the Phillips curve relationship flatter, suggesting smaller changes in wages are associated with bigger changes in unemployment when wage growth is low.

Krugman's data is annual, so I thought it would be interesting to look at the monthly relationship. Here's monthly unemployment with the monthly year-over-year change in wages:

What we see is that in fact the disinflationary/deflationary pressure of the recession did not simply cause us to slide down the Phillips curve during the recession. Instead we've actually spiraled around it, remaining well above the Phillips curve as unemployment was rising, and well below the curve now as unemployment is falling.

And finally, lets compare with the original Phillips curve from AW Phillips 1951 1958 paper:
This looks like Krugman's graph, albeit shifted downwards a bit. Now, like Krugman, Phillips looked at annual rather than monthly data. But there were plenty of depressions in his data similar to the Great Recession. I wonder if, had he plotted the monthly data, we'd see the same clockwise spiral around the phillips curve in those depressions.