teffects ipw PO means

[Adrienne Wold]

Hello,
I am estimating ATEs using teffects ipw, and I am wondering how to interpret the significance of the potential outcome means. In particular, what does it mean for the potential outcome mean to be statistically significant but not the ATE?

[Jeff Wooldridge]

The ATE is the difference in the two potential outcome means. So you’re not finding a statistically significant treatment effect. Rarely would the PO means themselves be different.

Per board etiquette, please show what you type and what Stata shows.