DiD control and treatment group size

[Lilly Tee]

Hi, I was wondering if there should be a concern about the sample size for the control and treatment groups. I am looking at the children's health following a policy that changes parental income. I found my total sample to be 2174. When I further consider breaking the sample into mothers who are 1,345 and fathers 829. When I consider the treatment, then the sample becomes the following:

Parent	Treatment = 0	Treatment = 1	Total
Father	773	56	829
Mother	1,257	88	1,345
Total	2,030	144	2,174

The samples are definitely not balanced. Would this be a problem since the sample size for the treated is very small? I would appreciate any thoughts

[Jared Greathouse]

DID is fine with relatively small treatment groups. No need to worry here (on its face anyways)

[Lilly Tee]

Ahh okay. Is there any reasoning for this? Because I do remember my lecturers saying in general don't have sample sizes of less than 50. Would recommend any additional readings on this