Event-study coefficients

Extract the event-study coefficients, standard errors, and confidence intervals from a fitted stacked regression — for custom plotting or combining models.

R stack_coefs()   ↔︎   Stata stored results (r(att), r(table), e()) after stacked reg

Usage

stack_coefs(model, conf_level = 0.95)

Arguments

Argument Description
model A model from stackreg() (or a fixest model on stacked data).
conf_level Confidence level for the intervals (default 0.95). 
stacked reg depvar, cluster(varname)

* the fitted coefficients live in e(b) / e(V); the parameter table in r(table)
matrix list e(b)
matrix list r(table)

* scalar ATT summaries
display r(att)      // average post-period ATT
display r(att_se)   // its delta-method SE

There is no separate extractor subcommand in Stata: after stacked reg the per-event-time coefficients are the interaction terms in e(b)/e(V) (with the standard r(table) layout), and the pooled summary is in the r() scalars.

R ↔︎ Stata mapping

R Stata
stack_coefs(model) e(b) / e(V) / r(table) after stacked reg
average post-period ATT (attr(model, "avg_post_att")) r(att), r(att_se), r(att_lb), r(att_ub), r(n_post)
per-cohort coefficients (stack_coefs() on a stackreg_groups object) r(group_att) after stacked reg, bygroup
conf_level level() on stacked reg

Value / Stored results

R returns a data.table with event_time, estimate, se, ci_lower, ci_upper. The reference period is included with estimate = 0, se = NA, and a ref_period attribute records which event time is the reference.

Stata: e(b)/e(V) hold the coefficient vector and variance; r(table) is the parameter table; r(att*)/r(n_post)/r(ref) hold the pooled ATT summary. After bygroup, r(group_att) has one row per cohort.

Example

library(stacked)
data(medicaid)
stack <- build_stack(medicaid, "year", "state", "adopt_year",
                     kappa_pre = 3, kappa_post = 2)
model <- stackreg(stack, "uninsured", cluster_var = "state")

stack_coefs(model)
   event_time     estimate          se     ci_lower     ci_upper
1:         -3 -0.001022172 0.003682642 -0.008240017  0.006195674
2:         -2 -0.003034560 0.002993349 -0.008901416  0.002832296
3:         -1  0.000000000          NA           NA           NA
4:          0 -0.016269503 0.003933884 -0.023979774 -0.008559231
5:          1 -0.023863697 0.006453761 -0.036512836 -0.011214558
6:          2 -0.025500057 0.007066243 -0.039349639 -0.011650476
stacked use medicaid, clear
stacked build, time(year) unit(state) adopt(adopt_year) kpre(3) kpost(2)
stacked reg uninsured, cluster(state)

matrix list r(table)
display "ATT = " r(att) "  (se " r(att_se) ")"

See also