Re: Difference-in-differences analysis (DiD)

[Sivathaasan Nadarajah]

Hi,
I am a PhD student and use Stata for my research work. I currently plan to do a difference-in-differences (DiD) analysis to solve the endogeneity issue arising from reverse causality. I have read few articles that use DiD approach to sort out reverse causality issue.

Particularly, those articles first use propensity score matching (PSM) strategy to construct a treatment and a control group.

Second, they sort the samples into terciles and retain the top and the bottom terciles.

Third, they employ PSM algorithm to identify matches between samples in the top and bottom tercile.

If possible, please see Table 5 in this article [Brogaard, J., Li, D., & Xia, Y. (2017). Stock Liquidity and default Risk. Journal of Financial Economics, 124(3), 486-502]

It would be grateful if you could assist me on this?
Thanks
Sivathaasan

[Red Owl]

The user written -diff- ado program (available on SSC) by Juan M. Villa can perform propensity score matching with diff-in-diff analyses. The help file for that user-written program contains useful examples.

ssc install diff

help diff

Hope that helps.

Red Owl
Stata/IC 15.1, Windows 10 (64-bit)

[Sivathaasan Nadarajah]

Thanks Red Owl for this help, which assisted me to do some practicals. Moreover, How do use propensity score matching to create a treatment and a control group. For the time-being, I have created these two groups based on median (e.g., treatment group =1 if it is above the median).