Tables 12 and 13 report our key results. Here we see the
o frictions" simulations when we
allow prices to adjust. We exclude the year 2006 because rms were likely experimenting, with
limited information about competitors' pricing strategies, in the rst year of the program and this
limits the validity of any comparison with the static pricing equilibrium. Consider rst the crossplan
unweighted average bids reported in columns 1 and 2 of Table 13. Recall that theory predicts
plans should respond to the removal of consumer inattention (i.e., increased search) by reducing
the rate at which they increase prices from year to year. The observed and simulated bids reported
in Table 13 for 2007-2009 are consistent with this intuition. The average simulated bid in 2007
is very similar to the average observed in the data for NJ PDP plans { both are essentially $80
per enrollee per month { while the median simulated bid is lower than the observed median ($77
compared to $81 observed in the data). Simulated bids then increase very little in 2008 and 2009,
approximately in line with the rate of cost increases, while the observed version has a much higher
growth rate.