Create industry experience dummy

[iangow]

For now, use two-digit SIC.

[iangow]

Hmm. Coverage is terrible.

Issues could be:

1. Too few activist directors matched to Equilar.
2. Too few company_id values matched to CRSP (using eight-digit CUSIPs).
3. Too few start_date or fy_end data values (for company_id).

library(dplyr, warn.conflicts = FALSE)
library(DBI)

pg <- dbConnect(RPostgres::Postgres())

rs <- dbExecute(pg, "SET search_path TO activist_director")

activist_directors <- tbl(pg, "activist_directors")
ind_exp <- tbl(pg, "ind_exp")

activist_directors %>%
    left_join(ind_exp) %>%
    count(prior_ind_exp)
#> Joining, by = c("campaign_id", "first_name", "last_name")
#> # Source:   lazy query [?? x 2]
#> # Database: postgres [igow@iangow.me:5432/crsp]
#>   prior_ind_exp       n
#>   <lgl>         <int64>
#> 1 FALSE             199
#> 2 NA               1500
#> 3 TRUE               59

^{Created on 2020-08-26 by the reprex package (v0.3.0)}

[sapyung]

I was able to match more with equilar as follows:

	_na	affiliated	unaffiliated	_na	affiliated	unaffiliated
0	16041	245	318	71%	75%	68%
1	5034	58	115	22%	18%	25%
2	1124	17	28	5%	5%	6%
3	304	4	4	1%	1%	1%
4	91	1	3	0%	0%	1%
5	33	1	1	0%	0%	0%
6	15	0	0	0%	0%	0%
7	12	0	0	0%	0%	0%
8	5	0	0	0%	0%	0%
9	1	0	0	0%	0%	0%
Total	22660	326	469	100%	100%	100%

However, I do not find a significant difference between activist directors and other directors.

@iangow Do you think we should give this another with BoardEx? Shall we include this in Table 3 as another column?

[sapyung]

I included this as Column 4 of Table 3 Panel A