In generalized DiD setting for staggered laws implementation, I am wondering if the treatment can be stopped in some lags rather than lasting the whole period.
The baseline identification:
$Y_{it}$ = $\alpha$ + $\beta$ $(Leniency Law)_{kt}$ + $\delta$$X_{ikt}$ + $\theta$$_t$ + >$\gamma$$_i$ +$\epsilon$$_{it}$ (1)
where $i$, $k$, and $t$ index firms, countries, and years respectively. $X_{ikt}$ is a vector of the different firm, country, and industry control, while $\gamma$ and $\theta$ are firm and year fixed effects.
For example, Dasgupta, 2019 examined the impact of anticollusion on firms' asset growth. If I argue that the impact of laws would be last around two years after the implementation, can I stop the serial observation of a country two years after its law implementation date?
For example, the sample period from 1990 to 2012, so if US passed the law in 1993, so from 1990 to 1992, $(Leniency Law)_{kt}$ will receive value of 0, then from 1993 to 1995, $(Leniency Law)_{kt}$ will receive the value of 1, and that's all for US data, so the sample for US data just from 1990 to 1995 rather than from 1990 to 2012 as the normal generalized DID setting.
Another sample is Korea passed the laws in 1997, so from 1990 to 1996, $(Leniency Law)_{kt}$ for Korean firms is 1 and $(Leniency Law)_{kt}$ for Korean firms in 1997-1999 is 1. Is it correct and acceptable?
