Shifting Points on a Graph/ normalising to 0 (Event Study Plot)

[Aaryan Deshpande]

Dear all,

I am currently writing my thesis on whether the 2007 indoor smoking ban had an effect on obesity in England. I am utilising a diff-in-diff strategy and aim to use an event study plot, below, to show that the parallel hypothesis holds.



I have a few questions regarding this graph:

1) How do I shift the graph downwards so that the point on the red line is on 0?
2) How do I move the red line along?
3) Can I claim that this plot suggests parallel trends?
4) Do I even need to shift this plot?

I used a template of commands and put my own variables in and therefore I am a bit confused. I put the following commands in to achieve the above plot:

reghdfe bmival2 sy* , vce(cluster hseyr gor) absorb( hseyr gor)

coefplot, keep(sy97 & sy98 & sy99 & sy00 & sy01 & sy02 & sy03 & sy04 & sy05 & sy07 & sy008 & sy09 & sy10 & sy11 & sy12 & sy13 & sy14 & sy15 & sy16 & sy17 & sy18) drop(_cons) xline(10) ciopts(recast(rcap)) recast(connected) vertical

Any help is hugely appreciated and not all questions have to be answered.

Thank you,
Aaryan