Event study revisions
(Updated 5/27/2021)

************************
1. Expanded time horizon:
************************
output in "output/eventstudy/extended/" performs the event study with an
expanded time horizon on the x axis. Before, it was -5 to 17, here it is -7 to
20 (ignore the first and last coefficient, they are binned with all earlier ages
and all later ages)

 (a) "z_composite_extended.xls"
     "z_dis_extended.xls"
     "z_ess_extended.xls"
     "z_hc_extended.xls"
     "z_nei_extended.xls"
     "dis_9_extended.xls"
     "ess_7_extended.xls"
are modifications from the event study code used to create 
Figure 3 and Figure 5 (use "all" sheet)
The event study coefficients in this table are "TBe_exp_ext_*"
Note that we ignore the first and last (TBe_exp_ext_1 and TBe_exp_ext_30)
coef's, these actually capture observations of many exposure levels (< -7, > 21)


  (b) the graphs in /coefplot/ are rough drafts of the event study figures, to
look at in the RDC instead of spreadsheets

For "all" sheet:
Column 5 = Model 0
Column 6 = Model 1
Column 7 = Model 2
Column 8 = Model 3

For "wm" "wf" "nm" "nf" sheet:
Column 3 = Model 1
Column 4 = Model 3

- Model 3 is the preferred model (includes survey year FE,
state x birth year FE, and county char trends) used in Figure 5 
- Model 2 is also used in figure 3 panel A
- plan to export the xls sheets and recreate figures
outside the RDC using the same post-processing code in the Dropbox

************************
2. 3 year bins
************************
output in "output/eventstudy/bins/" does the event study model with 3 year age
bins: [-5, -3], [-2, 0], [1, 3], [4,6], [7,9], [13,15], [16,18]
omitted: [10,12]

(a)  "z_composite_bins.xls"
     "z_dis_bins.xls"
     "z_ess_bins.xls"
     "z_hc_bins.xls"
     "z_nei_bins.xls"
     "dis_9_bins.xls"
     "ess_7_bins.xls"
 are modifications from the event study code
output used to create Figure 3 and Figure 5
The event study coefficients in this table are "TB_bin_*"
again, ignore first and last coefs (TB_bin_1 and TB_bin_10)

(b) the graphs in /coefplot/ are rough drafts of the event study figures

For "all" sheet:
Column 1 = Model 0
Column 2 = Model 1
Column 3 = Model 2
Column 4 = Model 3

For "wm" "wf" "nm" "nf" sheet:
Column 1 = Model 1
Column 2 = Model 3
