capture log close

use $"/$Users$/$abdul$/$Desktop$/$Presentation$/$njmin$.$dta", replace

describe

summarize

use $"/$Users$/$abdul$/$Desktop$/$Presentation$/$njmin$\_$reshaped.dta", replace

describe

summarize

$**$this is exploring the data to confirm the structure matches the description in the paper and codebook. $**$

use $"/$Users$/$abdul$/$Desktop$/$Presentation$/$njmin$.$dta", replace

$*$ Generate full$-$time equivalent $($FTE$)$ employment variables before and after the wage change

gen FTE$\_$BEFORE $=$ EMPFT $+$ NMGRS $+0\mathrm{.}5*$ EMPPT

gen FTE$\_$AFTER $=$ EMPFT$2+$ NMGRS$2+0\mathrm{.}5*$ EMPPT$2$

$*$ Create a variable for the gap in wages relative to the new minimum wage in NJ

gen GAP $=$ cond$($STATE $==1$ & WAGE$\_$ST $<5\mathrm{.}05,(5\mathrm{.}05-$ WAGE$\_$ST$)/$ WAGE$\_$ST$,0)$

$*$ Descriptive statistics: Mean and Standard Error for NJ and PA before and after the wage change

$*$ Compute means and standard deviations by state for before and after wage change

egen MEAN$\_$FTE$\_$BEFORE $=$ mean$($FTE$\_$BEFORE$),$ by$($STATE$)$

egen SD$\_$FTE$\_$BEFORE $=$ sd$($FTE$\_$BEFORE$),$ by$($STATE$)$

egen N$\_$BEFORE $=$ count$($FTE$\_$BEFORE$),$ by$($STATE$)$

gen SE$\_$FTE$\_$BEFORE $=$ SD$\_$FTE$\_$BEFORE $/$ sqrt$($N$\_$BEFORE$)$

egen MEAN$\_$FTE$\_$AFTER $=$ mean$($FTE$\_$AFTER$),$ by$($STATE$)$

egen SD$\_$FTE$\_$AFTER $=$ sd$($FTE$\_$AFTER$),$ by$($STATE$)$

egen N$\_$AFTER $=$ count$($FTE$\_$AFTER$),$ by$($STATE$)$

gen SE$\_$FTE$\_$AFTER $=$ SD$\_$FTE$\_$AFTER $/$ sqrt$($N$\_$AFTER$)$

$*$ List these values for a quick view

list STATE MEAN$\_$FTE$\_$BEFORE SE$\_$FTE$\_$BEFORE MEAN$\_$FTE$\_$AFTER SE$\_$FTE$\_$AFTER if $\_$n $==1|\_$n $==\_$N

$*$ Difference$-$in$-$Differences analysis

gen DELTA$\_$FTE $=$ FTE$\_$AFTER $-$ FTE$\_$BEFORE

gen TREATMENT $=($STATE $==1)$

$*$ Regression for difference$-$in$-$differences estimate

reg DELTA$\_$FTE TREATMENT

$*$ Check unique values in CHAIN to understand how many dummies are needed

tabulate CHAIN, missing

$*$Generate dummy variables for each chain if they are categorical and not numeric IDs

$*$ Let's create dummy variables for chain $2,3,$ and $4$ as an example

gen CHAIN$2=($CHAIN $==2)$

gen CHAIN$3=($CHAIN $==3)$

gen CHAIN$4=($CHAIN $==4)$

$*$ Run the regression with the newly created dummy variables

reg DELTA$\_$FTE TREATMENT CO$\_$OWNED CHAIN$2$ CHAIN$3$ CHAIN$4$

est tab

$*$To export the regression results to an external file

outreg$2$ using results.doc, replace word

$*$ Interaction terms and non$-$linear effects

gen WAGE$\_$SQ $=$ WAGE$\_$ST$^2$

gen INTERACTION $=$ TREATMENT $*$ WAGE$\_$ST

reg DELTA$\_$FTE TREATMENT WAGE$\_$ST WAGE$\_$SQ INTERACTION

est store model$1$

$*$ Check for robustness by adding$/$removing variables

reg DELTA$\_$FTE TREATMENT WAGE$\_$ST WAGE$\_$SQ

est store model$2$

reg DELTA$\_$FTE TREATMENT WAGE$\_$ST

est store model$3$

reg DELTA$\_$FTE TREATMENT CO$\_$OWNED CHAIN$2$ CHAIN$3$ CHAIN$4$ WAGE$\_$ST WAGE$\_$SQ

est store model$4$

$*$Displaying the Model in a comparative table

estout model$1$ model$2$ model$3$ model$4,$ cells$($b$($star fmt$(3))$ se$($par fmt$(3)))$ stats$($N r$2,$ fmt$(33)$ labels$("$Number of obs" $"$R$-$squared"$))$

gen LOW$\_$WAGE$\_$STORE $=($WAGE$\_$ST $<5\mathrm{.}05)$

$*$ Run models for the low wage stores

reg DELTA$\_$FTE TREATMENT if LOW$\_$WAGE$\_$STORE $==1$

est store low$\_$wage$\_$model

$*$ Table $7$ focus on different time effects or regions

reg DELTA$\_$FTE TREATMENT i$.$YEAR i$.$REGION

est store time$\_$region$\_$model

$*$ Check the first few entries of DATE$2$

list DATE$2$ in $1/10$

describe DATE$2$

$*$ Convert DATE$2$ to string and ensure it's zero$-$padded to $6$ characters

gen str$\_$date $=$ string$($DATE$2,"\%06\mathrm{.}0$f$")$

$*$ Extract month, day, and year components

gen MONTH $=$ real$($substr$($str$\_$date, $1,2))$

gen DAY $=$ real$($substr$($str$\_$date, $3,2))$

$*$ Adjust year for YY format to $1900$s or $2000$s

gen YEAR $=$ real$($substr$($str$\_$date, $5,2))$

replace YEAR $=$ YEAR $+1900$ if YEAR $<100$

$*$ Combine components into a Stata daily date

gen date $=$ mdy$($MONTH$,$ DAY, YEAR$)$

$*$ Now create YEAR and MONTH variables again for clarity

drop YEAR MONTH

gen YEAR $=$ year$($date$)$

gen MONTH $=$ month$($date$)$

$*$ Run regression with YEAR and MONTH as factors

reg DELTA$\_$FTE TREATMENT i$.$YEAR i$.$MONTH CHAIN$2$ CHAIN$3$ CHAIN$4$

est store monthly$\_$model

esttab monthly$\_$model

$*$ Run regression with YEAR and MONTH as factors

reg DELTA$\_$FTE TREATMENT i$.$YEAR i$.$MONTH CHAIN$2$ CHAIN$3$ CHAIN$4$

est store monthly$\_$model

$*$ Show results

esttab monthly$\_$model

$*$ Regression with the GAP variable

reg DELTA$\_$FTE GAP

$*$ Regression with GAP and additional controls

reg DELTA$\_$FTE GAP CO$\_$OWNED CHAIN$2$ CHAIN$3$ CHAIN$4$

$*$ Output results of regression models

estimates store Model$1$

estimates store Model$2$

estimates store Model$3$

estimates store Model$4$

estout Model$1$ Model$2$ Model$3$ Model$4,$ cells$($b$($star fmt$(3))$ se$($par fmt$(3)))$ stats$($N r$2,$ fmt$(33)$ labels$("$Number of obs" $"$R$-$squared"$))$

$*$ Regression model including chain ownership and indicator variables for chains

reg DELTA$\_$FTE TREATMENT CO$\_$OWNED CHAIN$2$ CHAIN$3$ CHAIN$4$

$*$ Include polynomial terms of continuous variables

gen WAGE$\_$SQ $=$ WAGE$\_$ST$^2$

$*$ Predict Cook's distance for the regression model

predict cooks$\_$d$,$ cooksd

$*$ Drop observations with a Cook's distance greater than $0\mathrm{.}5$

drop if cooks$\_$d $>0\mathrm{.}5$

gen TREATMENT$\_$ALT $=$ cond$($STATE $==1,1,0)$

which estout

ssc install estout, replace

ssc install outreg$2,$ replace

reg DELTA$\_$FTE TREATMENT

outreg$2$ using results.doc, replace

reg DELTA$\_$FTE TREATMENT CO$\_$OWNED CHAIN$2$ CHAIN$3$

Please help correct the STATA code to replicate CARD AND KRUEGER PAPAER ON Minimum wage.