Worksheet #3 Bacon

clear
local units = 3
local start = 1
local end	= 10

local time = `end' - `start' + 1
local obsv = `units' * `time'
set obs `obsv'

egen id	   = seq(), b(`time')  
egen t 	   = seq(), f(`start') t(`end') 	

sort  id t
xtset id t

lab var id "Panel variable"
lab var t  "Time  variable"

gen D = 0
replace D = 1 if id==2 & t>=5
replace D = 1 if id==3 & t>=8
lab var D "Treated"

gen Y = 0
replace Y = D * 2 if id==2 & t>=5
replace Y = D * 4 if id==3 & t>=8

lab var Y "Outcome variable"

* ##### Plotear Los resultados
set scheme plotplainblind 
twoway ///
	(connected Y t if id==1) ///
	(connected Y t if id==2) ///
	(connected Y t if id==3) ///
		,	///
		xline(4.5 7.5) ///
		xlabel(1(1)10) ///
		legend(order(1 "id=1" 2 "id=2" 3 "id=3"))		

* ### Correr Regresions 
estimates clear
eststo: xtreg Y D i.t, fe 
eststo: reghdfe Y D i.t, absorb(id)  
esttab, se

ssc install bacondecomp

* ### Decomposition
bacondecomp Y D, ddetail


* ### Ahora plotea un event-study. 
sort id t
bys id: egen g=min(t) if D==1
replace g=0 if D==0 
bys id: egen timing=max(g)
replace timing=11 if id==1

gen distance=t-timing 
gen distance_10=distance+10 

reg Y i.t i.id i.distance_10
coefplot, keep(*.distance_10) vertical recast(connected) xlabel(,angle(45)) xline(10) yline(0)

set seed 13082021
clear
local units = 30
local start = 1
local end   = 60

local time = `end' - `start' + 1
local obsv = `units' * `time'
set obs `obsv'

egen id	   = seq(), b(`time')  
egen t 	   = seq(), f(`start') t(`end') 	

sort  id t
xtset id t

lab var id "Panel variable"
lab var t  "Time  variable"

cap drop Y
cap drop D
cap drop cohort
cap drop effect
cap drop timing

gen Y 	   = 0		// outcome variable	
gen D 	   = 0		// intervention variable
gen cohort = .  	// total treatment variables
gen effect = .		// treatment effect size
gen timing = .		// when the treatment happens for each cohort

* Vamos a general un efecto que sea aleatorio entre 2 y 10
levelsof id, local(lvls)
foreach x of local lvls {
	local chrt = runiformint(0,5)	
	replace cohort = `chrt' if id==`x'
}

levelsof cohort , local(lvls)  //  let all cohorts be treated for now
foreach x of local lvls {
	
	// (a) effect
	
	local eff = runiformint(2,10)
		replace effect = `eff' if cohort==`x'
		
	// (b) timing	
	
	local timing = runiformint(`start' + 5,`end' - 5)	
	replace timing = `timing' if cohort==`x'
		replace D = 1 if cohort==`x' & t>= `timing' 
}

replace Y = id + t + cond(D==1, effect * (t - timing), 0)

/* El tiempo para cada cohorte también se genera aleatoriamente en el intervalo t=5 y t=55. Esto es solo para asegurarse de que las cohortes de tratamiento no sean muy dominantes, solo existen durante un par de períodos. */

levelsof cohort
local items = `r(r)'

local lines
levelsof id

forval x = 1/`r(r)' {
	
	qui summ cohort if id==`x'
	local color = `r(mean)' + 1
	* colorpalette tableau, nograph
		
	local lines `lines' (line Y t if id==`x', lc("`r(p`color')'") lw(vthin))	||
}

twoway ///
	`lines'	///
		,	legend(off)


* Corriendo la regresion

		estimates clear 
		eststo: xtreg Y i.t D, fe
		eststo: reghdfe Y D i.t, absorb(id)  
		esttab , keep(D) se
		* Pero sabemos que el tratamiento esta entre 2 y 10! Que paso?

		bacondecomp Y D, ddetail

* #### Event study

		gen distance= t-timinig
		gen distance_54=distance+54
		reg Y i.id i.t i.distance_54
		coefplot, keep(*.distance_54) vertical recast(connected) xlabel(,angle(45)) xline(54) yline(0)