Dear Stata users,

I want to estimate the ATE of a new policy on the number of firms, where policy has 4 treatments (0 not treated, treatment levels 1, 2 and 3). I am using teffects and inverse probability weights (ipw). I run the following command as indicated in the user guide

teffects ipw (number_firms pop2000 empl_rate lit_rate, logit) (multi)

where pop2000, empl_rate and lit_rate are covariates based on which I want to balance the model. Multi is a multivalued treatment variable. However, the outcome is an error saying that options are not allowed. Data are attached in the csv file.

I would appreciate your help on estimating multivalued treatment ipw model in teffects.

Thank you in advance,
Alina