1.

Table 1 in the AEJ-Macro paper has a typo for the forecast 
error variance at the 4 quarter horizon for the price markup 
shock. It should read:

                                    Output  Hours  Inflation

           6- Price markup shock      8.0    7.3     42.8


and then the first block of numbers adds up to 100.

2.

There is a mistake in the calculation of the standard deviation
of the markup shock in the CES case (in ./ces). In the paper
we said the the markup had a mean of 50 with a standard deviation
of 2,587%.  The corrected estimate is

           50 +/- 25.87 %

Note, however, that the range over the sample path is [-16,112]
and the period-by-period changes are large. Thus, we should
probably soften our conclusion by saying that the implied
markups are empirically implausible rather than saying the
standard deviation is absurdly large.  The codes on the website 
have been updated to reflect the correction.  My thanks to Peter 
Klenow for tracking down the precise place of the error.

Ellen McGrattan
June 2010